A Switching Observer for State-of-Charge Estimation of Reconfigurable Supercapacitors
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis work by authors is another view with ESR also represented in the circuit, but I have two major queries to be addressed in order for this manuscript to be published.
1. What is the effect of impedance with changing ESR and how does it effect SOC behavior?
2. Must include Kalman filtering model in this work and explain further.
Comments on the Quality of English LanguageEnglish seems to be alright.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsAuthors propose a new method for supercapacitor State-of-Charge (SOC) estimation. To demonstrate the benefits of the proposed method, the authors used graphical representations in which the estimation of SOC by different methods was shown. The Figures are of low quality with a very rough time scale and SOC scale (for example, Figures 4, 6, 8,9 and 10)). In Fig. 1. the voltage on the supercapacitor capacitance and the voltage on the supercapacitor terminals are written in capital letters, which is inconsistent labeling of physical quantities with the rest of the paper. In the paper, there are numerous errors in the expression of physical quantities (A detailed list is given below after the general comment). Graphs on some Figures are not correct, as well as provided comments and conclusions (a detailed description of the problem is given after the general comment). The paper did not describe in detail the measurement procedure, sensors positions types of sensors, which physical quantities are measured and what is the error (measurement uncertainty) of their measurements. The paper only provided a photograph (which is not described) of the hardware platform used for experiments, and a comment, I quote, "we conducted a series of experiments using a hardware platform similar to the one described in [19]" Overall, the paper is full of errors, omissions, incorrect graphics and incorrect statements. The results shown are not consistent with the theoretical prediction (a detailed comments for the above are given below after the general comment). The importance of SOC measurement in real life applications is poorly described in the paper. Paper failed to descibe shortcomings of measuring SOC using the open-loop estimation approach. Unfortunately, due to the above, I cannot recommend paper for publication.
Detailed comments to the authors
1. In the paper, there are numerous errors in the expression of physical quantities, for example:
Line 136, it is written 2.7V, it should be written 2.7 V, numeric value and units should be separated!
Line 136, it is written 2.2 m, it should be written 2.2 mOhm
Line 137, it is written 2A, it should be written 2 A
Line 238, it is written 370F, it should be written 370 F
Line 239 it is written 2mOhm, it should be written 2 mOhm
Line 239 it is written 2.7V, it should be written 2.7 V
Line 251 it is written 2A, it should be written 2 A
Line 252 it is written 0.1V, it should be written 0.1 V
Line 286, it is written 0.1V, it should be written 0.1 V
Line 286, it is written 0.9V, it should be written 0.9 V
Line 287, it is written 2A, it should be written 2 A
2. When describing the parameters on the basis of which the Figures were obtained, it is necessary to specify all relevant parameters. For example:
In Section 2.2.2. Limitations of Open-Loop SOC Estimation, It is stated that the supercapacitor has a rated voltage of 2.7 V and an internal resistance of 2.2 mOhm, and that charging takes place with a constant current of 2 A.
However, the supercapacitor's capacitance is not specified!
3. It is important to avoid sentences that can be interpreted in different ways.
For example:
Lines 247-249, I quote "To simulate real-world scenarios where component parameters may deviate from their nominal values, we introduce fluctuations of 10% and 20% in the system parameters. "
The dilemma remains, does the fluctuation of 10% and 20% in the system parameters refer to supercapacitor capacitance or supercapacitor series resistance (ESR), or both at the same time?
4. The established practice of writing papers is to refer to each individual Figure in the text, for example:
In section 4, entitled Experimental Validation, on Figure 6, individual graphic representations are marked with letters from a to c. However, in the text before Fig. 6. there are no individual references to the Fig. 6a, Fig. 6b and Fig. 6c. There is only general reference to Fig. 6.
5. Section 2.2.2. Limitations of Open-Loop SOC Estimation
In section 2.2.2, the authors claim that the parameter variation, specifically the capacitance of the supercapacitor, is the limiting factor in Open-Loop SOC Estimation. To confirm the claim, the authors used Figure 4. I quote "Figure 4 illustrates the SOC estimation results obtained using the open-loop method described by equation (12). To simulate real-world measurement uncertainties, we introduce a 10% fluctuation in capacitance (C). As evident from the figure, the estimated SOC fails to converge to the true SOC in the presence of parameter variations. "
The quoted claims are not valid!
The capacitance of the supercapacitor does not affect the accuracy of the SOCa measurement using the previous technique. This is indirectly stated in the paper (expressions (4) and (12)), SOC is independent of the capacitance, it depends on the ratio vc/vrated regardless of the supercapacitor capacitance. If SOC measurements are carried out with a current of 2 A (example from the paper), and if the supercapacitor capacitance is increased by 10%, due to the increased capacitance, charging until SOC=1 is reached will take longer, that is, the curve describing the SOC change will be less inclined, i.e. it will have a smaller slope. When the measurement is performed with constant current, the change in SOC is described by the expression:
SOC(t)=(Ich*t)/(C*Vrated) + SOC(t=0)
Where: t is the charging time, Ich is constant charging current, C – supercapacitor capacitance, Vrated is supercapacitor rated voltage, SOC(t=0) is inital SOC.
Comparing two supercapacitors that are charged with equal currents (in the paper 2 A), it is expected that the one with a higher capacitance has a lower SOC at a certain time. This clearly follows from the equation I gave. Therefore, the slope of the SOC curve is not an indicator of convergence, nor does it indicate an error.
The blue graph in Figure 4, which refers to a supercapacitor that has a capacitance variation of 10% (although the authors did not specify, since the blue graph has a smaller slope compared to the red graph, the capacitance is increased), should have been a straight line up to SOC=1. However, for an unknown reason, the charging of the supercapacitor with the higher capacitance was abruptly interrupted at the moment corresponding to the moment when the supercapacitor with the lower capacitance was charged. In this way, the reader may come to the wrong conclusion that if there is a capacitance variation, SOC can only be determined up to a certain limit. Also, the reader might come to the wrong conclusion that if there is a capacitance variation, there is a problem with convergence.
6. In section 4.1. "Experimental Results", the authors stated that the measurement is performed with a constant current of 2 A. A supercapacitor with the following parameters was used for the measurement: C=370 F, ESR 2 mOhm and Vrated=2.7 V.
In Figures 6, 7, 9 and 10, the authors presented graphs for SOC determined by different approaches. Figures also shows a blue graph, which according to the author's markings has the meaning, I quote "true SOC". Due to the very coarse division of the time scale, and the very coarse division of the SOC scale in Fig. , I had to zoom in on the Figure to extrapolate the values!
The validity of the blue graphs on the above Figures is disputed.
If the reader pays attention to the blue curve shown in Fig. 6a. (according to the author's markings, I quote "true SOC", the following can be observed (after a lot of zooming in the Figure 6a):
Initial SOC is approximately 0.04
The final SOC is approximately 0.35
Therefore, the change in SOC is approximately 0.31
Charging of the supercapacitor starts at 7.4 s
Charging of the supercapacitor ends at 122 s
Therefore, the duration of supercapacitor charging with a constant current of 2 A according to Fig. 6a (blue graph) is 114.6 s.
However, according to the theory, in order for the SOC to change in the range specified by the authors, a charging time (tch) is required which is determined by the expression:
tch=((SOCend-SOCinitial))*Vrated*C)/Ich
Inserting numerical values from paper gives charging time:
tch=(0.31 2.7*370)/2=154.8 s (approximately 155 s)
Therefore, there is a decrepancy of an incredible 40.2 s
Also, according to theory, SOC does not change in the range shown in Fig. 6a.
According to the theory, when charging of the supercapacitor is performed with a constant current, the change in SOC is determined by the expression
(SOCend-SOCinitial)=(Ich*tch)/(C*Vrated)
Inserting numerical values from paper gives SOC change:
(SOCend-SOCinitial)=(2*114.6)/(370*2.7)=229.2/999=0.229 (approximately 0.23)
According to the blue graph in Fig. 6.a SOC changed by 0.31, which is an error of 34.8 % in relation to the theoretical exact value.
All of the above can also be applied to Figures 7, 9 and 10. Because of the above, the conclusions presented in the paper are disputed.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have addressed the queries to the satisfaction. Hence I consider this work to be published in this journal without any further modifications.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors provide a detailed response to the comments made in the review. The authors have significantly improved the paper according to the comments in the review. The authors have corrected errors and omissions in the paper. The authors gave adequate answers regarding my doubts about the presented results. All my concerns about the paper have been removed.