The highway segments crossing small towns are defined as special road segments with the characteristics of both urban roads and highways. They are formed due to population and various production factors gathering on both sides of the highway. This type of road section is densely populated on both sides, and pedestrians, motor vehicles, and non-motorized vehicles interfere with each other, resulting in complex traffic flow conditions. At present, in small towns in China, the average number of electric bicycles per 100 households is 75.1. Electric bicycles have replaced bicycles as an important mode of travel for residents and become the main choice for short and medium distance travel [
1,
2]. On the highway segments crossing small towns, electric bicycles overspeed and occupy the motor carriageway, which makes the traffic conditions more diversified and the traffic flow more complicated and has a great impact on the safety and efficiency of traffic operation [
3]. At the same time, it also puts forward higher requirements for non-motorized traffic management and control strategies on the highway segments crossing small towns.
Due to the differences in the basic conditions of each country and the influence of the socio-economic level and city size, the population base and non-motorized vehicle ownership in each country are much lower than those in China [
4]. At the same time, many countries have legal restrictions on the movement of non-motorized vehicles on highways. For example, the New York Traffic Code prohibits the use of bicycles on highways, expressways, main roads, interstates, and overpasses unless there is a permit sign. In addition, the Canadian Federal Safety Act also prohibits electric bicycles from being allowed on highways, expressways, and other prohibited roads [
5]. Therefore, the conflict between motorized and non-motorized vehicles on the road is not as obvious as in China. Therefore, foreign scholars have focused their research on traffic safety, non-motorized riding behavior characteristics, and non-motorized traffic flow parameters. Den analyzed traditional bicycles and electric bicycles in terms of safety and listed the traffic safety hazards of electric bicycles in various aspects [
6]. Starting from the perspective of non-motorized vehicle riders, Taylor explored the maximum and optimal speed reduction of non-motorized vehicles under different traffic environments by establishing the individual gap acceptance probability independent decision model between motor vehicles and non-motor vehicles under the condition of mixed driving [
7]. Smith observed the riding conditions of conventional bicycles on the road and established the relationship model among the three parameters of traffic flow [
8]. Chinese scholars have more extensive research on the characteristics of non-motorized vehicle traffic. In terms of theoretical research on non-motor vehicle speed, flow, and density, Ping analyzed the basic characteristics of non-motorized vehicle traffic in the road section where motor vehicles and non-motorized vehicles drive separately and created the optimal relationship model of speed–density and speed–flow [
9]. Zeng studied the speed characteristics and density characteristics of non-motorized vehicles on urban roads. It is proved that the vehicles in the non-motorized lane affect each other, and their speed obeys normal distribution [
10]. Mao et al. studied the travel characteristics of electric bicycles using a logit regression model and concluded that the distance and time distribution of electric bicycle travels conformed to a quadratic curve model [
11]. Zhu et al. studied the relationship between the speed and density of non-motorized vehicles and proved that speed and density are exponential functions under congestion, and speed decreases with increasing density [
12]. Some scholars have modeled the travel distance of residents on the basis of analyzing the land use status. Chen et al. proposed to use the two-stage distribution model of Ireland to simulate the distance and spatial distribution state of total urban non-motorized travel [
13]. Qu et al. introduced the electron energy level theory into the non-motorized travel model of residents and used the simulated electron cloud resident travel distribution model as the basis to build a resident travel distribution model [
14].
In summary, existing studies focus on non-motorized traffic on urban roads, and few studies have been conducted on the non-motorized traffic characteristics of highway segments crossing small towns, so this area is a blind spot for research. The main feature of the highway crossing small towns in China is mixed non-motorized traffic, and the relevant research results have little significance for the study of the traffic characteristics of mixed non-motorized traffic [
15]. The large number of non-motorized traffic flow on highway segments crossing small towns has a great impact on the operation efficiency and safety of highways and puts forward higher requirements on the spatial and temporal planning and management of non-motorized riding [
16]. Based on this, the second part of this article studies the spatial and temporal distribution of non-motorized traffic on highway segments crossing small towns in the form of a survey questionnaire. Using this method, the patterns and characteristics of non-motorized traffic in terms of travel purpose, distance, time, and frequency can be obtained. The third part describes the research method of traffic flow parameters in four aspects: conceptual assumptions, sampling method, tool description, and analysis scheme. In the fourth part, based on the actual collected traffic data, the speed–density relationship and flow rate–density relationship of non-motorized traffic flow are studied using mathematical and statistical methods. Meanwhile, the proposed conceptual hypotheses are verified. The fifth part uses SPSS to perform regression analysis of the two variables, speed and distance, as linear, quadratic, and cubic functions. Based on the values of the coefficients of determination, the final chosen functional model is determined, and the correctness of the model is verified using relevant data from other cities.
It is very common to see motor vehicles and non-motorized vehicles mixed driving on highway sections through small towns. Therefore, it is very necessary to set up non-motorized lanes on the highway. However, the length of the non-motorized lane needs to be studied. Too long non-motorized lanes not only have a low road utilization rate but also will occupy too much road slab space and affect the normal movement of motor vehicle flow. The research results of this article can provide a theoretical basis for determining the setting range of non-motorized lanes in the segments of highways crossing small towns.