Research on the Spillover Effect of Different Types of Technological Innovation on New Energy Industry: Taking China’s Solar Photovoltaic as an Example

Abstract

Technological innovation has always played a very important role in the development of new energy industries. This paper takes the solar photovoltaic industry as an object of study, taking into account the diffusion of technological advances and the different roles of different technological innovations, and uses a spatial econometric SDM model to analyze the spillover effects of different types of technological advances on the solar industry in China. It is found that for the PV industry, efficiency-enhancing technological advances have the most significant impact, with efficiency-enhancing technologies contributing significantly to the annual electricity production of the PV industry; safety-enhancing technological advances having the second highest impact on the industry’s development; and cost-reducing technological advances have no significant impact on the industry. The study also found that due to the positive externalities of technological innovation, the spillover effect of technological innovation between regions has a significant impact on the development of the regional solar PV industry. In the long term, the direct effect of efficiency-enhancing technological innovation on the development of the PV industry is significantly positive, while the direct effect of safety-enhancing technological innovation on the development of the PV industry is significantly negative. Therefore, in the future, China’s solar energy industry should combine the capital investment of different types of science and technology into research and development, fully consider the impact of regional and technological spillover on industrial development, use technological innovation spillover to promote technological exchange and progress, and continuously improve the level of equipment operation safety, output efficiency, and electricity cost.

1. Introduction

In the process of energy transition, strengthening innovation in new energy technologies is an important means of efficient and high-quality use of clean energy. Many countries around the world have regarded energy technology as a breakthrough in the new round of scientific and technological revolution and industrial revolution, and major energy powers have formulated policies and measures to strengthen technological innovation in new energy fields such as solar energy. The United States has released the Inflation Reduction Act, the European Union has formulated the Renewable Energy Development Act, and Japan has introduced the Energy and Environment Innovation Strategy for 2030. All of them regard new energy technology as a breakthrough in the new round of scientific and technological revolution and industrial revolution and have formulated various policy measures to seize the high ground of development. The “Implementation Plan on Promoting the High-quality Development of New Energy in the New Era” issued by the National Energy Administration of China, the “14th Five-Year Plan” for Renewable Energy Development, the “14th Five-Year Plan for Modern Energy System”, and other plans all mention the need to embrace innovation and improve the level of renewable energy utilization. In addition, tax support tools, including a low 15% tax rate for high-tech enterprises and additional deductions for R&D expenses, have to some extent reduced the cost of R&D and promoted technological innovation for enterprises in the field of new energy such as solar energy. As the world’s second-largest economy and a leader in green economy technology, China has shown great power in undertaking green and low-carbon development of energy. Currently, among all new energy generation methods, solar power generation has a large scale and is an important part of the new energy industry. Solar energy utilization technologies mainly include photovoltaic power generation, solar thermal power generation, photochemistry, light induction, and photobiological conversion. The grid-connected capacity of photovoltaic power generation accounts for more than 99% of all installed solar power capacity, and photovoltaic power generation has become the mainstay of solar power generation technology. Faced with the new situation of achieving the “carbon peaking and carbon neutrality goals”, China’s solar energy utilization technology has ushered in a new period of development. At present, the manufacturing capacity of the solar power industry continues to strengthen, the industrial chain is gradually improved, the scale of development ranks first in the world, and the level of technology continues to improve but still faces some challenges.

First, in the field of research and development, China’s energy science and technology level is partially leading and partially advanced in the world, and there is still a gap between the overall quality and the international advanced level. Second, the cutting-edge technology of the solar energy industry has not yet been mastered, and some core equipment, processes, and materials are still subject to people; in particular, the trend of relying on imported equipment for major energy projects in the field of solar photovoltaics is still prominent. Third, the lack of time for technological innovation, application, and testing may bring stability and quality risks. For example, the research and development of large-capacity unit technology in China’s optoelectronic field, the technical application of new materials, energy storage, and the upgrading and innovation of production lines require a certain amount of time to verify, and the reduction rate of China’s photovoltaic electricity price level is too fast, and it is not coordinated with technological progress and cost decline speed, which will amplify the unit technology and quality risk [1]. Although the importance of technological innovation in promoting the development of the solar energy industry is undoubted, the mechanism of different types of technological innovation in China’s solar energy industry is not clear, and the spillover effect of technological innovation also significantly affects the layout and development of the new energy industry. Therefore, there is an urgent need to explore the mechanism of technological innovation in China’s solar energy industry, the spillover effect of technological innovation, and the impact of different types of technological innovation on the solar energy industry.

The importance of the spillover effect of technological innovation, one of the key factors affecting the new energy industry, to the development of the industry is self-evident.

Based on this, the spillover effect and the impact on the solar photovoltaic industry from the perspective of different types of technological innovation has become a frontier and a hot spot in the current research on innovation in the energy industry.

This paper compares the current state of development of China’s solar industry and the main challenges faced in the development of solar energy utilization technology and explores three different types of solar power utilization technology in view of the new situation of achieving the “carbon peaking and carbon neutrality goals” target and the characteristics of solar power generation. This paper aims to study the development of solar photovoltaic power generation for China’s solar industry, analyze the impact of technological innovation and economic, natural, and regional factors on the solar industry’s power generation, and use this analysis to derive countermeasures to meet China’s energy security needs in the context of a carbon-neutral future.

2. Literature Review

In the era of innovation-driven development, a new round of energy technology revolution is breeding and emerging, and new energy science and technology achievements are constantly emerging. Currently, scholars at home and abroad have conducted in-depth studies from different perspectives, including national, provincial, industrial, and enterprise perspectives around the evaluation of innovation capability [2,3], innovation efficiency [4], innovation transfer and diffusion [5,6,7], and influencing factors [8,9].

On the whole, first of all, the evaluation of the innovation ability of the new energy industry is considered. Research on the evaluation of innovation capacity is mainly carried out from a regional and sectoral perspective. Most research on technological innovation evaluation or measurement will first establish a corresponding index system. The evaluation system of innovation ability by domestic and foreign scholars can be mainly divided into a single index system and a comprehensive index system [10,11,12,13,14,15]. In the existing research, most scholars have chosen the number of patents as an indicator to measure the technological innovation ability of the new energy industry. Under normal circumstances, after foreign scholars establish the evaluation system and select the corresponding indicators, the next step will measure their weight. In terms of innovation ability evaluation methods, scholars at home and abroad mostly use traditional empirical evaluation, principal component analysis, arithmetic average weighting, factor analysis method, gray correlation analysis method, fuzzy mathematical method, CDA model, CCR model, and set matching method to analyze index data.

Second, the level of research related to the innovation efficiency of the new energy industry is considered. From the existing domestic and foreign research results, it can be seen that the existing research on technological innovation efficiency mainly uses two different measurement methods, namely the parametric method and non-parametric method, and carries out research from three different research levels at the enterprise, industry, and regional levels [16,17,18,19,20,21,22,23]. Among them, the representative of the parameter method is the SFA model. The representative of the non-parametric method is the DEA model, and the DEA model is also the research model adopted by most scholars at present; because it is difficult to obtain at the enterprise level, most of the research uses data above the scale or listed companies to represent the entire new energy industry. In addition to the research on the entire new energy industry, many scholars have also studied the innovation efficiency of the photovoltaic industry, wind power industry, hydropower industry, and other industries in the new energy industry.

Third, the factors influencing the technological innovation ability of new energy are considered. At home and abroad, there is more research on the influencing factors of innovation efficiency and regional innovation ability. Most studies have found that innovation input, innovation environment, industrial clusters, policy environment, etc., are the main factors affecting innovation [24,25,26]. Wang Hongwei (2022) made a corresponding assessment of the impact of the electricity price subsidy policy on the photovoltaic industry chain [27]. Wang Qunwei et al. (2013) systematically analyzed the constraints and interaction of technological innovation of new energy enterprises from four dimensions from the perspective of the technological innovation process [28]. In general, the research results on the influencing factors of new energy technology innovation have been very rich, but most of the studies do not consider the different types of technological innovation.

Fourth, the impact of the diffusion effect of technological innovation on the new energy industry cannot be ignored. The spillover effect of technological innovation has significantly promoted or inhibited the development of the new energy industry, which has attracted widespread attention from governments, enterprises, and scholars. At this stage, some scholars are conducting research from the perspective of the technology diffusion model. Kumar (2016) constructed a technology diffusion model for renewable energy [29]. Wang Hongying et al. (2022) further constructed a diffusion model of leading technologies in the new energy industry and analyzed the diffusion mechanism [30]. Some scholars have also explored the diffusion of technological innovation from the perspective of the spatial and temporal distribution of technological innovation. Xie Cong et al. used spatial econometric models and other methods to explore the spatiotemporal distribution and technology spillover of innovation in China’s new energy industry [31]. In general, there are relatively few studies exploring the spillover effect of technological innovation in the new energy industry, especially the solar photovoltaic industry, and the differences in the impact mechanism and degree of different technologies on the industry are not fully considered according to the characteristics of the new energy industry.

The literature in this field presents a large and fruitful body of research on technological innovation and the solar photovoltaic industry based on a specific technology or engineering project and rarely includes systematic and rigorous causal identification from the perspective of social science. Existing studies still need to be approached with caution. First, traditional panel regression usually assumes that the features of each region are independent of each other, which obviously deviates from reality, so it is necessary to break the traditional regression to introduce spatial panel econometric model analysis. Second, to what extent do technological innovation and other factors affect the solar photovoltaic industry, and does the spillover effect between regions have an impact on the solar photovoltaic industry in neighboring regions? Therefore, the spatial effects of different driving factors on the solar photovoltaic industry under the influence of spatial interaction between regions include direct effects and spillover (indirect) effects, which need to be studied in depth. Third, the impact of technological innovation on the solar photovoltaic industry cannot be generalized; different types of technological innovation will have different impacts on industrial development, and simply considering the impact of technological innovation level on industrial development obviously deviates from reality.

In view of this, compared with previous research, this paper mainly expands on the following aspects: (1) This paper uses the spatial econometric SDM model to analyze the temporal and spatial laws and influence mechanisms of technological innovation in the solar energy industry, and it takes into account the spatial spillover effect of technological innovation. (2) Different types of technological progress have different impacts on the mining solar energy industry; this paper comprehensively focuses on the characteristics of power generation cost, construction operation and equipment operation safety, power generation efficiency, etc., and combines the actual needs of the energy industry to divide solar energy technology progress into three categories.

3. Research Methodology and Data Selection

3.1. Spatial Measurement Global Model

Considering that the level of technological innovation interacts spatially, the flow of technology and patents between each other will have a certain contribution to the level of development of the solar industry in its neighboring provinces. Traditional econometrics does not take into account the spatial correlation between subjects, and the estimates may be biased [32,33]. This paper therefore considers the impact of technological innovation on the solar industry at a macro level using a spatial econometric global model. Different spatial econometric models assume different spatial transmission mechanisms, and their indicators represent different economic implications. In order to make the research process more scientific and rational, this paper starts with the spatial Durbin model (SDM), which is a nested form of spatial error model (SEM) and spatial lagged model (SLM), and its expressions are

$$\mathit{ln}{H}_{\mathit{it}}={\beta }_0+\theta W\mathit{ln}{H}_{\mathit{it}}+{\beta }_1\mathit{ln}{X}_{\mathit{core}}+{\beta }_2\mathit{ln}{X}_{\mathit{control}}+{\gamma }_{1}W\mathit{ln}{X}_{\mathit{core}}+{\gamma }_{2}Wln{X}_{\mathit{control}}+a_i+{\eta }_t+{\epsilon }_{\mathit{it}}$$

(1)

where $H_{\mathit{it}}$ is the explanatory variable, representing solar industry output;$X_{\mathit{core}}$ indicates the core explanatory variable; $X_{\mathit{control}}$ is the control variable, representing other influences related to solar industry output; ${\beta }_0$ is the constant term; $\theta$ is the spatial autoregressive coefficient; ${\epsilon }_{\mathit{it}}$ is the independent identically distributed random error term; $a_i$ and ${\eta }_t$ represent spatial and temporal fixed effects, respectively; ${\beta }_i$ and ${\gamma }_i$ correspond to the coefficients to be estimated for the spatial variables; and W represents the spatial weights.

3.2. Spatial Autocorrelation Test Model

The spatial autocorrelation test is mainly divided into the spatial correlation test and spatial heterogeneity test. The spatial correlation between the level of technological innovation and the level of development of the solar industry among regions is reflected in the trend of patent applications and the output of the solar industry between adjacent regions, which influence each other. The spatial heterogeneity of the level of technological innovation and solar industry development between regions is reflected in the fact that the number of patent applications and the supply of resources are not homogeneous in space. Therefore, this paper selects global and local correlation tests.

Global Moran’s I is used to test whether there is a spatial autocorrelation effect for socio-economic phenomena. Its formula is as follows:

$$\mathit{Moran}’s I=\frac{n{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}(x_i-\overline{x})(x_j-\overline{x})}{s^2{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}}$$

(2)

where $\overline{x}=\frac{1}{n}{\sum }_{i=1}^n{x}_i$, $s^2=\frac{\sum _{i=1}^n(x_i-\overline{x}{)}^2}{n}$, n denotes the number of spatial administrative regions in which the economic phenomenon is investigated, $x_i$ denotes the corresponding regional characteristic value, and $W_{ij}$ denotes the spatial weight, which measures the degree of spatial correlation in terms of geographical distance. If Moran’s I the variable is positively spatially correlated, the more it tends to 1, the stronger the correlation is; if Moran’s I takes a value between (−1, 0), the variable has a negative spatial correlation, and the closer it is to −1, the stronger the negative correlation is; if Moran’s I takes a value of 0, then the variable is randomly distributed and there is no spatial effect.

The global spatial autocorrelation of socio-economic phenomena is measured through the global Moran index. As economic phenomena are often spatially unbalanced, in order to reveal this unbalance, the local Moran’s I is used to present the spatial heterogeneity exactly. Its formula is as follows:

$$\mathit{Local} \mathit{Moran}’s I=\frac{(x_i-\overline{x})}{s^2}{\sum }_{j=1}^n{W}_{ij}(x_j-\overline{x})$$

(3)

Here, the indicators are the same as those in (2). The development of the solar industry can affect not only companies in the region but also companies in other regions through spatial spillover effects, thanks to the advancement of technology and economic development promotion, so this paper uses a spatial econometric model to examine the relationship between the two. The selection and setting of the spatial weight matrix are crucial to the analysis of spatial effects. The spatial weight matrices are as follows: (1) geographic adjacency weight matrix, where two regions sharing a common boundary are considered to be adjacent, with a weight of 1, otherwise 0; (2) geographic distance weight matrix, where the geographic distance weight is calculated based on the inverse of the Euclidean distance or radian distance, and as it involves a large range of scenarios, this paper uses the Earth’s radian curve distance calculation, where the closer the distance, the greater the weight; (3) economic distance weight, using the real GDP per capita measure. The geographical proximity weight matrix only considers whether regions directly border each other and does not fully consider the geographical correlation existing between provinces, while the economic distance weight matrix is constructed based on cross-sectional data, which can only examine the influence of exogenous factors such as spatial economic-social linkages on spatial effects and cannot accurately examine the transmission effect of geospatial effects, and the relationship between wind power generation and one of the key factors influencing the location and efficiency of wind and photovoltaic power generation projects is the natural influence factor, which can be reflected more accurately by different geographical locations. Therefore, geographical distance between regions plays a crucial role in the development of the industry, its spillover, and the diffusion of the level of technological progress. In order to fully examine the impact of geographical location on solar energy companies, a geographical distance weighting matrix is used for the analysis.

3.3. Selection of Indicators and Data Sources

The development of the solar industry is influenced by many complex factors such as the natural environment, policy guidelines, economic development, and technological innovation [34,35,36,37]. In selecting the influencing factors, representative influencing factors were selected based on previous studies and screened according to the degree of influence on the solar industry and the ease of obtaining information [38,39,40]. The research by Chen Fengnan and others shows that government policies, technological levels, and consumer markets are the main factors affecting the spatial pattern of China’s solar photovoltaic industry. The studies of Huang Xiang and others have found that different climatic conditions and solar radiation data are significantly related to photovoltaic power generation. Wang Jiaming and others point out that technological conditions and socio-economic levels are important factors influencing the development of China’s photovoltaic industry.

According to the current situation of the development of the solar photovoltaic industry, the four influencing factors of resource endowment, socio-economic development, technology, and natural environment were selected for analysis, taking into account the characteristics of the solar photovoltaic industry and the availability of data [1,41,42,43,44,45,46,47].

Resource endowment: Two indicators were selected to express the abundance of solar resources in terms of average annual solar light intensity and the stability of solar resources in terms of solar sunshine hours [26].

Socio-economic factors: The determination of how to quantitatively analyze the role of various influencing factors on the development of the photovoltaic power generation industry and transform important influencing factors into quantifiable economic indicators is an important element of the systematic analysis of the development of the photovoltaic power generation industry. Combined with China’s current technical conditions, the rapid and healthy development of the photovoltaic power generation industry depends on whether the level of socio-economic development can support the rapid expansion of the photovoltaic power generation industry, in addition to increasing the government’s macro-control efforts and adjusting international trade strategies. In this paper, the GDP of each region is selected to indicate the level of local socio-economic development.

Technological factors: Technological progress has an important influence on industrial development [32], especially in the photovoltaic industry as a new energy source. This paper analyzes technological progress as a key influencing factor for the development of the solar industry and classifies technological progress into three types according to the characteristics of the solar PV industry: efficiency-enhancing, safety-enhancing, and cost-reducing. The number of solar PV technology patents is chosen as the explanatory variable for the technology factor.

The natural environment: Finding the key factors that determine the overall development speed strategy of wind power and photovoltaic power generation is of great significance for the rational formulation of the development strategy of the solar industry and the stable development of society, the economy, and the environment, and also has a certain guiding value for investors in the solar market. Research on the micro level of photovoltaic power generation has been relatively mature, with most research scholars taking a particular project or region or even a single device as the object of study and exploring in detail the impact of different natural factors on the solar industry’s power generation. Based on the insights and conclusions of previous research scholars, factors such as cloudiness, light intensity, light duration, humidity, and temperature were selected as control variables.

In this paper, sample data from 31 provinces and municipalities directly under the Central Government of China from 2010 to 2020 were selected to analyze the impact of technological innovation and other factors on the development of the solar industry, with cloudiness, light intensity, light duration, humidity, and temperature selected as natural factor control variables, GDP selected as a social factor control variable, and annual electricity production of the solar industry selected as the explanatory variable. Due to the impact of the 2020 epidemic, which resulted in distorted data, the 2020 data were excluded from the spatial econometric modeling phase of this paper. The specific results are shown in

Table 1. Indicator system.

Solar Industry	Index
The variable being explained	The solar photovoltaic industry generates electricity annually by province
Explanatory variables	Three types of technological innovations
Control variables	Natural factors: cloud cover, light intensity, light duration, humidity, temperature; social factors: GDP

.

Explanatory variables: Solar and wind power generation, the representative industries of solar energy, were selected as the object of this paper, and the annual power generation of each region in China was used to represent the measurement of the development of the solar industry in each region.

Core explanatory variables: The number of patent applications by region was selected as a proxy for technological innovation variables, i.e., as the core explanatory variable in this paper.

Control variables: Cloudiness, light intensity, light duration, humidity, and temperature were selected as natural factor control variables, and GDP was selected as a social factor control variable.

The data on technological innovation were obtained from patent databases such as Incopat, and the data on annual industry electricity production and GDP were obtained from the China Statistics Bureau. The rest of the control variables were taken from the International Energy Agency (IEA) website and the China Meteorological Administration website.

4. Spatial Autocorrelation Test

4.1. Global Spatial Autocorrelation Test

With the help of GeoDa software, the global Moran’s I index is used to discuss the correlation characteristics of the level of technological innovation and the spatial distribution of electricity production in the solar industry. The results are shown in

Table 2. Global Moran’s I index of electricity generation in the photovoltaic industry.

Year	I	E(I)	sd(I)	z	p-Value *
2011	0.052	−0.033	0.030	2.857	0.002
2012	0.056	−0.033	0.026	3.456	0.000
2013	0.141	−0.033	0.028	6.113	0.000
2014	0.125	−0.033	0.032	4.950	0.000
2015	0.109	−0.033	0.033	4.274	0.000
2016	0.078	−0.033	0.034	3.310	0.000
2017	0.055	−0.033	0.035	2.550	0.005
2018	0.062	−0.033	0.035	2.708	0.003
2019	0.054	−0.033	0.035	2.478	0.007
2020	0.089	−0.027	0.071	1.71	0.05

* indicates the significance of the result.

and

Table 3. Global Moran’s I index of the level of technological innovation.

Year	I	E(I)	sd(I)	z	p-Value *
2011	0.021	−0.033	0.025	2.186	0.014
2012	0.040	−0.033	0.025	3.002	0.001
2013	0.141	−0.033	0.028	6.113	0.000
2014	0.125	−0.033	0.032	4.950	0.000
2015	0.109	−0.033	0.033	4.274	0.000
2016	0.078	−0.033	0.034	3.310	0.000
2017	0.055	−0.033	0.035	2.550	0.005
2018	0.053	−0.033	0.027	3.162	0.001
2019	−0.072	−0.033	0.022	−1.734	0.041
2020	0.021	−0.033	0.025	2.186	0.014

* indicates the significance of the result.

. The tests show that the global Moran’s I index values for solar industry output are both positive and negative and most of them pass the 10% significance test. The Moran’s I indices for the years that passed the test ranged from 0.057 to 0.095, indicating that there is a spatial concentration of regions with similar solar industry output. The global Moran’s I index for technological innovation is positive, and the Moran’s I index for 2010–2021 ranges from 0.133–0.187, and all pass the 10% significance test, indicating a strong spatial clustering trend in the level of technological innovation. Therefore, the level of technological innovation has a strong spatial aggregation effect, and there is a positive spatial correlation with solar industry output.

4.2. Local Spatial Autocorrelation Test

This paper further explores the local spatial autocorrelation of the level of technological progress and solar industry electricity generation, using the local Moran’s I test and the GeoDa software to obtain the following scales and distribution maps, respectively, and this paper selects the results of the local spatial autocorrelation test between the level of technological innovation in 2020 and solar industry output in 2020 for analysis. The results are shown in Figure 1, Figure 2, Figure 3 and Figure 4 and

Table 4. Moran’s Ii (2020—PV production).

Province	Ii	E(Ii)	sd(Ii)	z	p-Value *
Anhui	0.155	−0.033	0.172	1.094	0.137
Beijing	−0.040	−0.033	0.237	−0.028	0.489
Chongqing	0.107	−0.033	0.119	1.176	0.120
Fujian	−0.095	−0.033	0.105	−0.583	0.280
Gansu	0.169	−0.033	0.165	1.230	0.109
Guangdong	0.033	−0.033	0.104	0.638	0.262
Guangxi	0.213	−0.033	0.126	1.960	0.025
Guizhou	0.132	−0.033	0.113	1.465	0.071
Hainan	0.195	−0.033	0.134	1.702	0.044
Hebei	−0.325	−0.033	0.159	−1.836	0.033
Heilongjiang	−0.090	−0.033	0.096	−0.596	0.276
Henan	−0.044	−0.033	0.196	−0.054	0.478
Hubei	−0.004	−0.033	0.121	0.240	0.405
Hunan	0.047	−0.033	0.113	0.705	0.240
Jiangsu	0.260	−0.033	0.180	1.632	0.051
Jiangxi	0.005	−0.033	0.128	0.295	0.384
Jilin	−0.039	−0.033	0.197	−0.029	0.488
Liaoning	−0.008	−0.033	0.141	0.182	0.428
Inner Mongolia	0.012	−0.033	0.118	0.385	0.350
Ningxia	0.129	−0.033	0.109	1.492	0.068
Qinghai	0.160	−0.033	0.174	1.107	0.134
Shaanxi	0.028	−0.033	0.120	0.510	0.305
Shandong	0.290	−0.033	0.182	1.776	0.038
Shanghai	−0.378	−0.033	0.143	−2.399	0.008
Shanxi	−0.027	−0.033	0.083	0.079	0.469
Sichuan	0.080	−0.033	0.122	0.927	0.177
Tianjin	0.034	−0.033	0.235	0.286	0.388
Xinjiang	0.014	−0.033	0.056	0.847	0.198
Tibet	−0.032	−0.033	0.064	0.016	0.493
Yunnan	0.099	−0.033	0.107	1.240	0.108
Zhejiang	0.238	−0.033	0.180	1.508	0.066

* One-tail test.

. The results show that technological innovation and solar power generation in some provinces show aggregation effects; for example, the level of technological innovation in the eastern coastal provinces shows high–high aggregation characteristics of local aggregation, and solar power generation in the northern part of the city shows high–high aggregation and in the southern part of the city shows low–low local aggregation characteristics. All the local aggregation regions passed the significance test, but most of them did not show local aggregation or autocorrelation.

5. Empirical Analysis of the Photovoltaic Power Industry

5.1. Regression Results without Consideration of Spatial Characteristics

Firstly, a geographically weighted spatial weight matrix is chosen to set up an OLS model for ordinary regression analysis, without considering spatial effects and the cumulative effect of technological progress, the results are shown in

Table 5. OLS regression.

Linear Regression
Solar Energy	Coef.	St. Err.	t-Value	p-Value	[95%Conf. Interval]		Sig
Cloud cover	0.792	0.137	5.79	0	0.522	1.061	***
Duration of daylight	0.354	0.129	2.74	0.007	0.1	0.609	***
Humidity	−0.767	0.108	−7.11	0	−0.979	−0.554	***
Temperature	0.258	0.115	2.25	0.025	0.032	0.484	**
GDP	0.251	0.089	2.83	0.005	0.076	0.426	***
Technological advances in safety	3.057	0.815	3.75	0	1.453	4.661	***
Cost-based technological advances	−2.924	0.999	−2.93	0.004	−4.89	−0.957	***
Efficiency-oriented technological advances	0.607	1.405	0.43	0.666	−2.16	3.374
Light intensity	0.36	0.121	2.96	0.003	0.121	0.599	***
Constant	−0.475	0.202	−2.34	0.02	−0.873	−0.076	**
Mean dependent var	0.131		SD dependent var			0.220
R-squared	0.420		Number of obs			279
F-test	21.617		Prob > F			0.000
Akaike crit. (AIC)	−185.451		Bayesian crit. (BIC)			−149.139

*** p < 0.01, ** p < 0.05.

. Immediately afterward, an F-test is conducted, using both the classical LM test proposed by Anselin and the robust LM test proposed by Anselin and Smirnov, the results are shown in

Table 6. LM test.

Test	Statistic	df	p-Value
Spatial error
Moran’s I	5.422	1	0.000
Lagrange multiplier	18.747	1	0.000
Robust Lagrange multiplier	23.748	1	0.000
Spatial lag
Lagrange multiplier	1.878	1	0.171
Robust Lagrange multiplier	6.880	1	0.009

. By determining the spatial correlation of the lag or residual terms, both the LM test for spatial errors and the robust LM test rejected the original hypothesis at the 1% level of significance. The LM test for spatial lags failed the test, but the robust LM test for spatial lags rejected the original hypothesis at the 1% level of significance, which would indicate a spatial correlation between the variables, and therefore the spatial error model was considered more appropriate. The joint significance of spatial and temporal fixed effects is also examined, and the results of the likelihood ratio test confirm that the model should be extended with both temporal and spatial fixed effects.

Based on the results in the table below, we find that the core explanatory variable safety-enhancing patents and the control variables cloudiness, light hours, temperature, GDP, and light intensity all have a significant positive effect on the solar industry’s photovoltaic power generation, but there are also a few types of technological advances such as cost-based patents and the control variable humidity that have a negative effect on the solar industry’s photovoltaic power generation.

5.2. Spatial Correlation Studies

In this section, we follow LeSarge and Pace’s recommendation that even if the non-spatial model is rejected in favor of using SEM or SAR based on the LM test, the Hausman, LR test is still required. The Hausman test results indicate the presence of fixed effects, and the use of a fixed effect model should be considered, with preference given to the SDM model, the results are shown in

Table 7. Hausman test.

Solar Energy	Coef.	Std. Err.	z	p > z	[95%Conf. Interval]
Main
Cloud cover	0.556	0.166	3.350	0.001	0.231	0.882
Duration of daylight	0.228	0.143	1.590	0.111	−0.053	0.509
Humidity	−0.628	0.133	−4.710	0.000	−0.889	−0.367
Temperature	0.225	0.127	1.770	0.077	−0.024	0.474
GDP	0.119	0.102	1.170	0.242	−0.081	0.319
Technological advances in safety	−1.618	0.938	−1.720	0.085	−3.456	0.221
Cost-based technological advances	−0.716	0.888	−0.810	0.420	−2.456	1.024
Efficiency-oriented technological advances	2.880	1.413	2.040	0.041	0.111	0 5.648
Light intensity	0.235	0.129	1.810	0.070	−0.019	0.488
_cons	0.038	0.903	0.040	0.966	−1.731	1.808
Wx
Cloud cover	−0.102	0.511	−0.200	0.842	−1.103	0.899
Duration of daylight	−0.434	0.518	−0.840	0.402	−1.449	0.580
Humidity	0.048	0.584	0.080	0.935	−1.098	1.193
Temperature	−0.293	0.735	−0.400	0.691	−1.733	1.148
GDP	−0.183	0.669	−0.270	0.784	−1.494	1.127
Technological advances in safety	−9.441	4.832	−1.950	0.051	−18.912	0.030
Cost-based technological advances	−3.501	2.676	−1.310	0.191	−8.745	1.744
Efficiency-oriented technological advances	16.136	6.623	2.440	0.015	3.156	29.116
Light intensity	0.132	0.668	0.200	0.843	−1.177	1.442
Spatial
rho	0.401	0.140	2.870	0.004	0.127	0.675
Variance
lgt_theta	0.332	0.338	0.980	0.325	−0.329	0.994
sigma2_e	0.013	0.001	10.990	0.000	0.011	0.016

.

This article conducts model goodness-of-fit comparisons through two types of LR tests and selects the most suitable research model and perspective for this study. The purpose of the first LR test is to compare the goodness of fit of time fixed effect models, space fixed effect models, and dual fixed effect models in order to choose the appropriate fixed effect model for this study, the results are shown in

Table 8. LR test—determining the type of fixed effect.

Likelihood-Ratio Test	LR chi2(9) = 107.21
Assumption: TIME Nested in SDM	Prob > chi2 = 0.0000
Likelihood-Ratio Test	LR chi2(9) = 20.50
Assumption: IND Nested in SDM	Prob > chi2 = 0.0151

. The purpose of the second LR test is to compare the SEM, SAR, and SDM models to determine which is most suitable for this research, the results are shown in

Table 9. LR test—determining whether the SDM model degenerates into an SEM vs. SAR model.

Likelihood-Ratio Test	LR chi2(9) = 32.14
Assumption: SAR Nested in SDM	Prob > chi2 = 0.0002
Likelihood-Ratio Test	LR chi2(9) = 35.14
Assumption: SEM Nested in SDM	Prob > chi2 = 0.0001

. The results of the LR test showed that the original hypothesis that the time fixed effect was superior to the double fixed effect model was significantly overturned, and the original hypothesis that the individual fixed effect model was superior to the double fixed effect model was significantly overturned. Therefore, in this paper, the LR test was conducted using the SDM model under double fixed effects as a sample to determine whether the SDM model is applicable to this study, and the results are as follows.

The Hausman and LR tests showed that the double fixed SDM regression model was more applicable to this study than the SAR to SEM regression model, and the LR tests for both SAR and SEM significantly rejected the original hypothesis, and the SDM model did not degenerate into a simple SEM, SAR model.

5.3. Empirical Analysis of Model Regression Results Considering Spatiotemporal Characteristics

Based on the SDM model, we used a time fixed effect regression model to derive the results of our study. The results are shown in

Table 10. SDM time fixed effect regression model.

Solar Energy	Coef.	Std. Err.	z	p > z	[95%Conf. Interval]
Main
Cloud cover	0.674	0.133	5.080	0.000	0.414	0.934
Duration of daylight	0.358	0.111	3.210	0.001	0.140	0.577
Humidity	−0.729	0.113	−6.450	0.000	−0.950	−0.507
Temperature	0.341	0.094	3.640	0.000	0.157	0.525
GDP	0.121	0.081	1.490	0.137	−0.038	0.281
Technological advances in safety	−1.727	0.949	−1.820	0.069	−3.588	0.133
Cost-based technological advances	−0.957	0.868	−1.100	0.270	−2.657	0.743
Efficiency-oriented technological advances	3.222	1.489	2.160	0.030	0.304	6.140
Light intensity	0.263	0.093	2.820	0.005	0.080	0.446

, it can be seen that the variables cloud volume, hours of light, humidity, temperature, and light intensity passed the 1% significance test, the variable efficiency-enhancing patents passed the 5% significance test, the variable safety-enhancing patents passed the 10% significance test, and the variables GDP and cost-reducing patents did not pass the significance test. This indicates that, from a macro perspective, light hours, humidity, temperature, efficiency-enhancing technological progress, cloudiness, safety-enhancing technological progress, and light intensity are important factors affecting the development of China’s solar industry.

For the core explanatory variables, the estimated coefficient of safety-enhancing patents is negative. For every 1% increase in the number of safety-enhancing patents, the annual electricity production of the photovoltaic industry decreases by 1.727%. The reason why an increase in safety-enhancing patents can significantly reduce the annual electricity production of the photovoltaic industry is that the rapid development of technology and its long history have led to a certain level of industrialization in the current solar industry. In contrast, people are more concerned about maximizing the efficiency and minimizing the cost of energy generation, and the investment in industrial safety has outweighed the benefits it brings, which does not play a corresponding role in promoting the generation of energy in the industry. The estimated coefficient of efficiency-enhancing patents is positive. For every 1% increase in the number of efficiency-enhancing patents, the annual power generation capacity of the photovoltaic power generation industry rises by 3.222%, and the increase in efficiency-enhancing patents can significantly increase the annual power generation capacity of the photovoltaic power generation industry because in the solar energy industry, efficiency-enhancing patents serve the speed of energy generation, and the increase in the number of efficiency-enhancing patents can, to a certain extent, reflect the increase in the efficiency of the industry. The impact of cost-reduction patents on the development of the solar industry is not significant, probably due to the delayed effect of technological advances on the development and planning of the industry and the fact that cost-reduction technological innovations are not immediate for the development of the industry. At the same time, even with the increasing level of cost-reducing technological innovation, the cost of electricity generated by solar power is still higher than the cost of conventional power generation. Under the influence of this current situation, the market’s subjective willingness to invest in the solar development industry will not increase significantly in the near future, and it is understandable that cost-reducing patents do not have a significant impact on the development of the solar industry.

For the control variables, the estimated coefficients of light duration, light intensity, and temperature are positive, which indicates that increases in light duration, light intensity, and temperature all result in a significant increase in the annual electricity production of the PV power industry. An increase in the duration of sunlight leads to an extended period for photovoltaic power generation equipment to absorb light, while an increase in light intensity results in more light being absorbed by the equipment per unit of time. The increase in sunlight duration extends the power generation time of photovoltaic devices, directly promoting their electricity output. The increase in light intensity boosts the power generation efficiency of photovoltaic devices per unit of time, directly promoting their electricity output within a fixed period. The increase in temperature is, to some extent, caused by the increase in light intensity, and the increase in light intensity increases the rate at which photovoltaic power generation equipment can absorb light in a fixed period of time, and both the longer time for absorbing light and the increase in efficiency result in a significant increase in the annual electricity production of the photovoltaic industry. The estimated coefficient of cloudiness is positive, which indicates that an increase in cloudiness results in a significant increase in the annual electricity production of the photovoltaic industry due to the following reasons: In recent years, when the temperature is too high, some parts of the power generation equipment may age faster, thus reducing the efficiency of power generation. The increase in the amount of clouds in the air can effectively reduce the excessive radiation of light, so that the surface temperature of the power generation equipment will not be too high, thus ensuring the efficiency of the equipment. The estimated coefficient of GDP is not significant; GDP is an important measure of national economic development, and the increase in GDP represents the progress of China’s economic level and technological development but may not have a more direct or significant impact on the development and progress of the PV industry. The estimated coefficient of humidity is negative, which indicates that an increase in humidity will cause a significant decrease in the annual power generation capacity of the PV industry. This is because an increase in air humidity will to some extent weaken the radiative effect of light, thus reducing the intensity of light, making the rate of light absorption by the PV industry during a fixed period of time decrease, and ultimately reducing the annual electricity production of the PV industry.

In terms of the coefficients of the spatial lag term, the W* independent variable indicates the impact of independent variables from other regions around a region on the annual electricity production of that region. The results are shown in

Table 11. Regression results with spatially lagged terms.

Solar Energy	Coef.	Std. Err. z		p > z	[95%Conf. Interval]
Wx
Cloud cover	2.548	0.990	2.570	0.010	0.608	4.487
Duration of daylight	2.394	1.118	2.140	0.032	0.201	4.586
Humidity	−1.028	0.754	−1.360	0.172	−2.505	0.449
Temperature	1.483	0.915	1.620	0.105	−0.309	3.276
GDP	−0.123	0.634	−0.190	0.846	−1.366	1.120
Technological advances in safety	−8.117	5.573	−1.460	0.145	−19.039	2.805
Cost-based technological advances	−4.628	5.011	−0.920	0.356	−14.449	5.194
Efficiency-oriented technological advances	15.732	8.847	1.780	0.075	−1.607	33.072
Light intensity	0.881	0.858	1.030	0.304	−0.800	2.562
Spatial
rho	0.241	0.194	1.240	0.215	−0.140	0.622
Variance
sigma2_e	0.016	0.001	11.780	0.000	0.013	0.019

, w* cloud amount passes the 5% significance test, corresponding to a positive estimated coefficient, which indicates that an increase in cloud amount in a neighboring province will have a significant positive impact on the annual electricity production of the PV power generation industry in this province. This indicates that an increase in the number of hours of light in neighboring provinces has a significant positive impact on the annual electricity production of the PV power generation industry in the province. The efficiency-enhancement patents passed the 10% significance test and the corresponding estimated coefficients were positive, indicating that an increase in the number of efficiency-enhancement patents in neighboring provinces would have a significant positive impact on the annual power generation capacity of the PV power generation industry in this province, and when the number of efficiency-enhancement patents in neighboring provinces increased by 1%, the annual power generation capacity of the PV power generation industry in this province increased by 15.732%.

Based on the SDM model, the results of the study were obtained separately using individual fixed effect regression models, and it can be seen that none of the variables passed the significance test except for the variables cloudiness, humidity, and GDP. Based on the SDM model, the results of the study were obtained separately using a double fixed effect regression model, and it can be seen that none of the variables passed the significance test except for the variables cloudiness and GDP and the three types of technological progress. The results are detailed in Appendix A.

5.4. Decomposition of Spatial Effects

In terms of direct effects, all variables except cost-reducing patents and GDP passed the significance test, and the positive and negative estimated coefficients of all variables were consistent with the estimated coefficients of the spatial lag model. Among them, the direct effect of efficiency-enhancing patents was the largest, followed by safety-enhancing patents, with humidity ranking third, followed by cloudiness, light duration, temperature, and light intensity, indicating that the core explanatory variables, efficiency-enhancing patents and safety-enhancing patents, all had a greater impact on the annual power generation of the PV industry, while the various control variables all had a smaller impact on the annual power generation of the PV industry. In addition, the absolute value of the estimated coefficient of efficiency-enhancing patents is greater than that of the estimated coefficient of the spatial lag model, because in the test, the direct effect takes into account the spatial feedback effect; i.e., the change in the number of efficiency-enhancing patents in the province affects the electricity generation of the PV industry in the province by affecting the electricity generation of the PV industry in the neighboring provinces. This feedback effect is partly from the spatially lagged dependent variable and partly from the spatially posterior independent variable. The results are shown in

Table 12. Direct effect results.

Solar Energy	Coef.	Std. Err. z		p > z	[95%Conf. Interval]
LR_Direct
Cloud cover	0.728	0.148	4.930	0.000	0.439	1.018
Duration of daylight	0.403	0.129	3.120	0.002	0.149	0.656
Humidity	−0.748	0.121	−6.180	0.000	−0.986	−0.511
Temperature	0.366	0.102	3.600	0.000	0.167	0.566
GDP	0.123	0.084	1.470	0.141	−0.041	0.287
Technological advances in safety	−1.833	1.039	−1.760	0.078	−3.869	0.204
Cost-based technological advances	−1.036	0.941	−1.100	0.271	−2.880	0.808
Efficiency-oriented technological advances	3.453	1.659	2.080	0.037	0.201	6.705
Light intensity	0.289	0.097	2.980	0.003	0.099	0.479

.

In terms of indirect effects, only the variables cloudiness and hours of light passed the 10% significance test, while all other variables failed the significance test; i.e., the spillover effects of all variables except for the variables cloudiness and hours of light had insignificant effects on the electricity production of the PV industry in other provinces. The results are shown in

Table 13. Indirect effect results.

Solar Energy	Coef.	Std. Err. z		p > z	[95%Conf. Interval]
LR_Indirect
Cloud cover	3.777	1.946	1.940	0.052	−0.037	7.592
Duration of daylight	3.461	2.037	1.700	0.089	−0.531	7.453
Humidity	−1.652	1.279	−1.290	0.197	−4.160	0.855
Temperature	2.122	1.511	1.400	0.160	−0.839	5.083
GDP	−0.059	0.899	−0.070	0.947	−1.821	1.702
Technological advances in safety	−11.770	9.261	−1.270	0.204	−29.922	6.381
Cost-based technological advances	−6.538	7.709	−0.850	0.396	−21.648	8.572
Efficiency-oriented technological advances	22.631	15.559	1.450	0.146	−7.864	53.126
Light intensity	1.323	1.279	1.030	0.301	−1.184	3.830

.

In terms of the total effect, only cloud amount passed the 5% significance test, and the two variables of light hours and humidity passed the 10% significance test. The positive and negative estimated coefficients of all variables were consistent with the estimated coefficients of the spatial lag model. Among them, the total effect of cloud amount is the largest, followed by light hours and humidity, which indicates that all three control variables have a significant positive effect on electricity generation in the PV industry. The results are shown in

Table 14. Total effect results.

Solar Energy	Coef.	Std. Err. z		P > z	[95%Conf. Interval]
LR_Total
Cloud cover	4.506	1.999	2.250	0.024	0.589	8.423
Duration of daylight	3.864	2.118	1.820	0.068	−0.287	8.014
Humidity	−2.401	1.332	−1.800	0.071	−5.010	0.209
Temperature	2.488	1.571	1.580	0.113	−0.590	5.567
GDP	0.064	0.944	0.070	0.946	−1.786	1.914
Technological advances in safety	−13.603	9.963	−1.370	0.172	−33.129	5.923
Cost-based technological advances	−7.574	8.077	−0.940	0.348	−23.405	8.257
Efficiency-oriented technological advances	26.084	16.733	1.560	0.119	−6.712	58.880
Light intensity	1.612	1.308	1.230	0.218	−0.952	4.176

.

5.5. Robustness Tests

Based on the fact that the study is on the relationship between the PV power generation industry and technological progress, the certainty that the PV power generation industry is affected by technological progress, i.e., the robustness of the model, needs to be further verified. In this paper, the level of technological progress in the total PV power generation sector is selected as the instrumental variable, and the same model is used again for modeling analysis using efficiency-enhancing, cost-reducing, and safety-enhancing technological progress in order to explore the robustness of the model results.

(1)

Replacement of efficiency-enhancing patent variables

The results of

Table 15. Robustness test results.

	Replacement Efficiency		Replacement Cost		Replacement Safety
Solar Energy	Coef.	p > z	Coef.	p > z	Coef.	p > z
Main
Cloud cover	0.647	0.000	0.645	0.000	0.670	0.000
Duration of daylight	0.377	0.001	0.341	0.002	0.379	0.001
Humidity	−0.652	0.000	−0.693	0.000	−0.681	0.000
Temperature	0.299	0.001	0.329	0.000	0.299	0.001
GDP	0.113	0.157	0.134	0.090	0.132	0.102
Technological advances in safety	−0.000	0.008	−0.000	0.005	−0.000	0.032
Cost-based technological advances	−0.509	0.388	−2.040	0.028	0.330	0.720
Efficiency-oriented technological advances	1.187	0.059	2.869	0.005	0.283	0.758
Light intensity	0.296	0.001	0.287	0.002	0.290	0.002
Wx
Cloud cover	2.515	0.011	2.377	0.015	2.669	0.007
Duration of daylight	2.278	0.041	2.151	0.053	2.354	0.035
Humidity	−0.349	0.643	−0.656	0.355	−0.598	0.437
Temperature	0.831	0.357	1.200	0.182	0.852	0.349
GDP	0.130	0.838	0.100	0.872	0.301	0.635
Technological advances in safety	−0.001	0.022	−0.001	0.024	−0.001	0.074
Cost-based technological advances	−2.892	0.395	−11.305	0.045	3.493	0.541
Efficiency-oriented technological advances	6.219	0.105	15.683	0.013	−0.750	0.893
Light intensity	1.393	0.108	1.281	0.132	1.193	0.169
Spatial
rho Variance	0.229	0.243	0.202	0.312	0.225	0.253
sigma2_e	0.016	0.000	0.016	0.000	0.016	0.000

were obtained by replacing the variables to bring the total level of technological progress in the field of photovoltaic power generation into the model. As can be seen from the model output, the significance conclusions obtained corroborate the previous time fixed effect SDM model, and the positive and negative values of the coefficients of each variable are also generally consistent with the previous time fixed effect SDM model, thus proving the accuracy of the conclusions already obtained.

(2)

Replacement of cost-reducing patent variables

The results in Table 15 were obtained by replacing the variables to bring the total level of technological progress in the PV power generation sector into the model. As can be seen from the model output, the significance conclusions obtained corroborate the previous time fixed effect SDM model, and the positive and negative values of the coefficients of each variable are also generally consistent with the previous time fixed effect SDM model, thus proving the accuracy of the conclusions already obtained. A comparison of the robustness results shows that efficiency-enhancing technological advances are still positively correlated and have the highest coefficient of influence on the PV industry overall, thus demonstrating the key influence of efficiency-enhancing technological advances on the PV industry.

(3)

Replacement of safety-enhancing patent variables

The results of Table 15 were obtained by replacing the variables in the model with the total level of technological progress in the PV power generation sector. It was found that after replacing the original variables in the model, the safety-enhancing technological progress, the replaced instrumental variables were highly significant while the other two technological progress factors became insignificant, indicating that the impact of cost-reducing and efficiency-enhancing technological progress on electricity generation in the PV industry is consistent with the total level of technological progress in the PV power generation sector. This indicates that the impact of cost-reducing and efficiency-enhancing technological advances on electricity generation in the PV industry is consistent with the overall level of technological advances in the PV industry.

6. Research Findings

This paper uses the input–output indicators of China’s solar industry over several decades to conduct a comprehensive evaluation and optimization modeling analysis of the factors influencing the power generation of China’s solar industry (photovoltaic) based on the spatial Durbin model, and it provides an in-depth discussion of the efficiency of the solar industry’s technological impact in the field of power generation and the variability of the impact of different types of technological advances on power generation; on this basis, it analyzes the optimization path of technological advances in the solar industry. Based on the above analysis and modeling process, the following main conclusions are obtained.

First, for the photovoltaic industry, the impact of efficiency-enhancing technological advances is the most significant, with efficiency-enhancing technological advances significantly boosting the industry’s annual electricity production, followed by safety-enhancing technological advances for the industry’s development, and the impact of cost-reducing technological advances on the industry’s development is not significant. Therefore, this paper argues that the impact of technological advances on the development of the solar industry is crucial, with particular attention being paid to the impact of efficiency-enhancing and safety-enhancing technological advances at this stage.

Second, the spillover effect needs to be taken into account in the study of the impact of technological advances on the development of the solar industry, but the diffusion effect of different types of technological advances and different solar industries varies. The study found that efficiency-enhancing technological advances have a significant spillover effect on the photovoltaic power generation industry, and the promotion effect on the power generation of the photovoltaic industry is very significant. Therefore, this paper indicates that the impact of technological advances on the development of the solar industry cannot be generalized and that differential analysis should be carried out for different industries and different types of technological advances, so as to tailor the development plan of the industry to the material and local conditions; at the same time, this paper finds that the PV power generation industry has a strong autocorrelation in terms of geographical location and has a strong spillover effect on the industrial layout in local areas, showing a low–low aggregation or high–high aggregation pattern. The paper also finds that the photovoltaic power generation industry has a strong autocorrelation in terms of geographical location and has a strong spillover effect on the industrial layout in local areas, mostly showing the local aggregation characteristics of low–low or high–high aggregation.

The research fully considers the natural, social, and technological factors that influence the development of the renewable energy photovoltaic industry and takes into account the spatiotemporal differences in the development of the renewable energy industry. The study focuses on the impact of different types of technological innovation on the development of the photovoltaic power generation industry, which will have a positive effect on the future development of China’s renewable energy industry. However, there is still room for improvement and further exploration in the research; for example, policy factors will significantly affect the development of the renewable energy industry in the early and middle stages, and national policy support will play a crucial guiding role. However, due to the difficulties in data collection and quantification of policy factors, this article does not delve into the impact of support policies on the development of the renewable energy industry and the promotion of policy on the level of industrial technological innovation. These issues need to be further studied in the future.

7. Policy Recommendations

The first recommendation is to strengthen investment support for R&D in solar energy technology innovation and to strengthen technology R&D in areas such as solar energy utilization, high-ratio solar energy grid connection and transmission, energy storage, and energy internet, focusing on the most promising technology themes such as battery storage technology, solar photovoltaic technology, and solar fuel technology.

The second recommendation is to establish a mechanism for tackling key core technologies in the solar energy industry in line with China’s national conditions; targeting key common technologies urgently needed by the industry, such as advanced energy storage, high-efficiency solar energy utilization, large wind turbine manufacturing, and carbon capture utilization and sequestration; actively guiding and supporting the joint participation of the scientific research community and industry; exploring the establishment of an industry-wide R&D platform led and organized by industry-leading enterprises; and carrying out joint innovation in solar energy technology.

The third recommendation is to strengthen the application and promotion of new technologies and pilot demonstrations and promote model innovation. The layout and construction of a technology innovation center in the field of solar energy should be planned to provide a platform for the application and promotion of new technologies and testing and verification for the development of the industry. Based on technologies such as high-efficiency solar power generation, high-precision power prediction on long time scales, intelligent dispatch control, and new energy storage, a number of solar green power plants with high reliability and flexibility will be built on a pilot basis to reduce the risk of new technology application.

Finally, because the layout of China’s solar industry planning only has a few decades of history, and the development of industrial scale and relatively high speed, it is inevitable that the layout and promotion of the industry will encounter many problems, and the future of China’s solar industry should be guided by policy, combined with different types of science and technology of capital investment in research and development, and constantly improve the level of safety in the use of equipment, operational safety, output efficiency, cost of electricity, etc. The industry will gradually shift from “winning by quantity” to a parallel track of “quality” and “quantity”.