Controllable load adjustable potential evaluation method considering user response will

Abstract

The invention discloses a controllable load adjustable potential evaluation method considering user response willingness, and the method comprises the steps: carrying out the quantitative aggregation of the user response willingness according to different power utilization characteristics of a flexible load, and constructing a flexible load influence factor membership function through a TSK fuzzy model; the relation between the up-regulation and down-regulation response willingness of the flexible load cluster and other influence factors is quantified, the upper and lower limits of the adjustable potential of the flexible load cluster are calculated according to the determined response willingness threshold value to serve as one of consideration of the reliability of the adjusting quantity, richer information is provided for optimized operation of the power distribution network, and the reliability of the power distribution network is improved. And a real-time electricity price scheme is combined to guide the electricity consumption behavior of the user, thereby reducing the power grid load peak and filling up the power grid load valley.

Description

Technical Field

[0001] The present invention relates to the technical field of load evaluation, and in particular to a method for evaluating the adjustable potential of a controllable load taking into account a user's response willingness. Background Art

[0002] With the access of large-scale distributed renewable energy, the uncertainty factors of distribution network now present the characteristics of wide sources, diverse categories and complex principles, which run through the entire distribution network operation process. Based on the analysis of a large amount of historical data, the uncertainty problems of distribution network can be divided into three categories: source, network and load, which correspond to the uncertainty problems of intermittent renewable energy generation fluctuations in distribution network, state fluctuations affected by important equipment and complex network structure, and load and demand side response fluctuations.

[0003] According to the data, among the existing methods for evaluating the potential of controllable load adjustment, there are cluster analysis such as K-means, hierarchical clustering and self-organizing neural network clustering to analyze the potential of load adjustment; there are methods for extracting the characteristics of electricity consumption time series components of SAE network to evaluate the response potential of controllable resources at multiple time points; and there are methods for simulating user demand response according to factors such as season and call time to evaluate the potential.

[0004] Chinese patent document CN115330280B discloses a "method and system for evaluating the adjustable potential of air-conditioning load demand response in aggregated loads". The method includes the following steps: establishing an air-conditioning load total amount evaluation model, evaluating the air-conditioning load total amount, and obtaining the air-conditioning load total amount curve; establishing a first-order equivalent thermal parameter model to obtain a mechanism model of a single air-conditioned room; based on the mechanism model of a single air-conditioned room, establishing a double-layer grouped rotation-controlled air-conditioning cluster structure and an air-conditioning cluster control strategy; based on the double-layer grouped rotation-controlled air-conditioning cluster control strategy, conducting simulation experiments on the air-conditioning cluster to obtain simulation results; based on the simulation results, obtaining a controllable proportional coefficient of the air-conditioning load; calculating the product of the total air-conditioning load and the controllable proportional coefficient of the air-conditioning load to obtain the actual controllable air-conditioning load demand response amount in the aggregated load. The above technical solution lacks consideration of the user's willingness to respond. Summary of the invention

[0005] The present invention mainly solves the technical problem that the original technical solution lacks consideration of user response willingness, and provides a controllable load adjustable potential evaluation method considering user response willingness, quantifies and aggregates user response willingness according to different power consumption characteristics of flexible loads, constructs the membership function of flexible load influencing factors using the TSK fuzzy model, quantifies the relationship between the upward and downward response willingness of the flexible load cluster and other influencing factors, and calculates the upper and lower limits of the adjustable potential of the flexible load cluster as one of the considerations for the reliability of the regulation amount, providing more abundant information for the optimized operation of the distribution network, and guiding user electricity consumption behavior in combination with the real-time electricity price scheme, thereby reducing the peak load of the power grid and filling the trough load of the power grid.

[0006] The above technical problem of the present invention is mainly solved by the following technical solution: The present invention comprises the following steps: S1 analyzing the influencing factors of typical flexible load on the demand side; S2 performs demand response modeling for typical flexible loads on the demand side; S3 uses TSK fuzzy model to establish the membership function of flexible load influencing factors; S4 quantifies the coupling between the upward and downward response willingness of the flexible load cluster and other influencing factors; S5 sets the response willingness threshold and calculates the upper and lower limits of the adjustable potential of the flexible load cluster.

[0007] According to the different electricity consumption characteristics of flexible loads, the response willingness of aggregated users is quantified, and the membership function of flexible load influencing factors is constructed using the TSK fuzzy model. The relationship between the upward and downward response willingness of the flexible load cluster and other influencing factors is quantified. According to the determined response willingness threshold, the upper and lower limits of the adjustable potential of the flexible load cluster are calculated as one of the considerations for the reliability of the regulation quantity, providing more abundant information for the optimized operation of the distribution network, and guiding users' electricity consumption behavior in combination with the real-time electricity price plan, thereby reducing the peak load of the power grid and filling the trough load of the power grid.

[0008] Preferably, the influencing factors of step S1 include the SOC value of the electric vehicle battery, the charging fee of the charging station, and the demand-side response adjustment amount. Taking the electricity price of the charging station as an example, when the electricity price is less than 0.2 yuan / kWh, the user will think that the electricity price is cheap, when the electricity price is between 0.4 and 0.6 yuan / kWh, the user will think that the electricity price is appropriate, and when the electricity price is higher than 0.8 yuan / kWh, the user will think that the electricity price is expensive. Taking the SOC of the electric vehicle battery as an example, when the SOC remaining amount of the electric vehicle battery is 30%, the user will think that the SOC remaining amount of the electric vehicle battery is insufficient; when the SOC remaining amount of the electric vehicle battery is 50%-70%, the user will think that the SOC remaining amount of the electric vehicle battery is appropriate; when the SOC remaining amount of the electric vehicle battery exceeds 80%, the user will think that the SOC remaining amount of the electric vehicle battery is sufficient.

[0009] As a preferred method, based on the real data of electric vehicles, the adjustable capacity and adjustment willingness of electric vehicles are analyzed, and the demand response modeling of typical flexible loads on the demand side is carried out. The cluster analysis method is used to analyze the historical load data of users and the specific electricity consumption patterns of different industries. On this basis, evaluation indicators for quantifying the response potential of various types of users are proposed, which has good adjustment potential for loads in the metal processing industry and tea processing industry; considering the timing response characteristics under the uncertainty of adjustable resources, it is helpful to obtain the probability distribution of the response potential of adjustable resource clusters at different time nodes, and to assist in the hierarchical and classified regulation of multi-type adjustable resources in the region. However, the above methods do not consider the inherent relationship between the user's response willingness and the electricity consumption pattern and electricity consumption time, and lack the specific quantification of the user's response willingness. Therefore, the user's response willingness can be quantified and aggregated according to the different electricity consumption characteristics of flexible loads, and the user's electricity consumption behavior can be guided in combination with the real-time electricity price scheme, thereby reducing the peak load of the power grid and filling the trough of the power grid load.

[0010] Preferably, the step S3 specifically includes: S3.1 defines corresponding fuzzy subsets for the three influencing factors of charging station electricity price, SOC and regulation amount; S3.2 obtains the dispatch intention under a single rule for each group of determined input electric vehicle clusters’ electricity price, battery SOC and regulation amount.

[0011] Preferably, the fuzzy rules corresponding to the TSK fuzzy model in step S3.2 are: Where: x1, x2 and x3 are the input variables charging station electricity price, electric vehicle battery SOC and regulation amount respectively; is the fuzzy subset corresponding to the input electricity price of the kth rule, is the fuzzy subset corresponding to the input variable SOC of the kth rule, is the fuzzy subset corresponding to the input variable adjustment amount of the kth rule; and is the fuzzy connection operator; f k (x) is the linear function value corresponding to the kth fuzzy rule; is the fixed parameter of the kth fuzzy rule; K is the number of fuzzy rules.

[0012] Preferably, the step S4 specifically includes: S4.1 When a single influencing factor remains unchanged, the relationship between the response willingness of the electric vehicle cluster and other factors is constructed; S4.2 Calculate the response intention of electric vehicles and air-conditioning clusters that are finally output by the TSK fuzzy model based on the influencing factors.

[0013] Preferably, the step S4.2 specifically includes: Where: and are the membership functions corresponding to the fuzzy subsets; EV is the response willingness value of the electric vehicle cluster.

[0014] Preferably, the step S5 specifically includes selecting a threshold value of a downward response willingness and an upward response willingness, and calculating an upper and lower limit of the adjustable potential of the flexible load cluster.

[0015] Preferably, the charging station electricity price x1 sets a judgment threshold. When the electricity price is less than 0.2 yuan / kWh, the user will think the electricity price is cheap. When the electricity price is between 0.4-0.6 yuan / kWh, the user will think the electricity price is appropriate. When the electricity price is higher than 0.8 yuan / kWh, the user will think the electricity price is expensive.

[0016] Preferably, the smaller the adjustment amount is, the higher the response willingness of the electric vehicle user is, and the higher the reliability is in the actual demand response process.

[0017] The beneficial effects of the present invention are: quantifying the coupling relationship between the flexible load response willingness and other influencing factors, being able to quantify the upper and lower limits of the load adjustable potential under different flexible load response willingness, as one of the considerations of the reliability of the regulation quantity, and providing more abundant information for the optimized operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flow chart of the present invention.

[0019] Figure 2 This is a typical distribution diagram of charging load of an electric vehicle cluster according to the present invention.

[0020] Figure 3 It is a schematic diagram of a membership function of a charging station electricity price according to the present invention.

[0021] Figure 4 It is a schematic diagram of an adjustable power lower limit with response willingness of 0.8 and 0.5 according to the present invention.

[0022] Figure 5 It is an adjustable potential result diagram of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the technical solution of the present invention is further described in detail below through embodiments and in combination with the accompanying drawings. It should be understood that the specific implementation method described here is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0024] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations (or steps) therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the drawings; the process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0025] Among the existing methods for evaluating the potential of controllable load adjustment, some use cluster analysis such as K-means, hierarchical clustering and self-organizing neural network clustering to analyze the load adjustment potential; some use the power consumption time series component feature extraction of the SAE network to evaluate the controllable resource response potential at multiple time points; and some simulate user demand response according to factors such as season and call time to evaluate the potential.

[0026] The cluster analysis method is used to analyze the historical load data of users and the specific electricity consumption patterns of different industries. On this basis, evaluation indicators for quantifying the response potential of various types of users are proposed, which has good regulation potential for the loads of metal processing and tea processing industries; considering the time series response characteristics under the uncertainty of adjustable resources, it is helpful to obtain the probability distribution of the response potential of adjustable resource clusters at different time nodes, and to facilitate the hierarchical and classified regulation of multi-type adjustable resources in the region. However, the above methods do not consider the inherent relationship between user response willingness and electricity consumption patterns and electricity consumption time, and lack specific quantification of user response willingness.

[0027] Therefore, the present invention provides a method for evaluating the adjustable potential of controllable loads taking into account user response willingness, which can quantify and aggregate user response willingness according to different power consumption characteristics of flexible loads, and guide user power consumption behavior in combination with real-time electricity price schemes, thereby reducing the peak load of the power grid and filling the trough load of the power grid.

[0028] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0029] Embodiment: This embodiment provides a method for evaluating the adjustable potential of controllable loads taking into account the user's response willingness, such as Figure 1 Shown, including Follow these steps: S1 analyzes the factors affecting typical flexible loads on the demand side. The influencing factors include the SOC value of the electric vehicle battery, the charging fee of the charging station, and the demand side response adjustment amount. Taking the charging station electricity price as an example, when the electricity price is less than 0.2 yuan / kWh, the user will think that the electricity price is cheap, when the electricity price is between 0.4 and 0.6 yuan / kWh, the user will think that the electricity price is appropriate, and when the electricity price is higher than 0.8 yuan / kWh, the user will think that the electricity price is expensive. Taking the electric vehicle battery SOC as an example, when the electric vehicle battery SOC remaining is 30%, the user will think that the electric vehicle battery SOC margin is insufficient; when the electric vehicle battery SOC remaining is 50%-70%, the user will think that the electric vehicle battery SOC margin is appropriate; when the electric vehicle battery SOC remaining exceeds 80%, the user will think that the electric vehicle battery SOC margin is sufficient.

[0030] S2 performs demand response modeling for typical flexible loads on the demand side. Based on the real data of electric vehicles, the adjustable capacity and adjustment willingness of electric vehicles are analyzed, and demand response modeling is performed for typical flexible loads on the demand side. Cluster analysis is used to analyze the historical load data of users and the specific electricity consumption patterns of different industries. On this basis, evaluation indicators for quantifying the response potential of various types of users are proposed, which has good adjustment potential for loads in the metal processing industry and tea processing industry; considering the timing response characteristics under the uncertainty of adjustable resources, it is helpful to obtain the probability distribution of the response potential of adjustable resource clusters at different time nodes, and to assist in the hierarchical and classified regulation of multi-type adjustable resources in the region. However, the above methods do not consider the inherent relationship between user response willingness and electricity consumption patterns and electricity consumption time, and lack specific quantification of user response willingness. Therefore, the user response willingness can be quantified and aggregated according to the different electricity consumption characteristics of flexible loads, and the user's electricity consumption behavior can be guided in combination with the real-time electricity price scheme, thereby reducing the peak load of the power grid and filling the trough of the power grid load.

[0031] S3 uses the TSK fuzzy model to establish the membership function of the factors affecting flexible loads, including: S3.1 defines corresponding fuzzy subsets for the three influencing factors of charging station electricity price, SOC and regulation amount; S3.2 For each group of determined input electric vehicle cluster electricity price, battery SOC and adjustment amount, the dispatch intention under a single rule is obtained. The fuzzy rules corresponding to the TSK fuzzy model are: Where: x1, x2 and x3 are the input variables charging station electricity price, electric vehicle battery SOC and regulation amount respectively; is the fuzzy subset corresponding to the input electricity price of the kth rule, is the fuzzy subset corresponding to the input variable SOC of the kth rule, is the fuzzy subset corresponding to the input variable adjustment amount of the kth rule; and is the fuzzy connection operator; fk (x) is the linear function value corresponding to the kth fuzzy rule; is the fixed parameter of the kth fuzzy rule; K is the number of fuzzy rules.

[0032] S4 quantifies the coupling between the upward and downward response willingness of the flexible load cluster and other influencing factors. Specifically, it includes: S4.1 When a single influencing factor remains unchanged, the relationship between the response willingness of the electric vehicle cluster and other factors is constructed; S4.2 Calculate the response intention of the electric vehicle and air conditioning cluster output by the TSK fuzzy model based on the influencing factors, including: Where: and are the membership functions corresponding to the fuzzy subsets; EV is the response willingness value of the electric vehicle cluster.

[0033] S5 sets the response willingness threshold and calculates the upper and lower limits of the adjustable potential of the flexible load cluster. Select the lower and upper response willingness thresholds and calculate the upper and lower limits of the adjustable potential of the flexible load cluster.

[0034] The charging station electricity price x1 sets a judgment threshold. When the electricity price is less than 0.2 yuan / kWh, the user is set to think that the electricity price is cheap. When the electricity price is between 0.4 and 0.6 yuan / kWh, the user will think that the electricity price is appropriate. When the electricity price is higher than 0.8 yuan / kWh, the user will think that the electricity price is expensive. Taking the charging station electricity price x1 as an example, when the electricity price is less than 0.2 yuan / kWh, the user is set to think that the electricity price is cheap. When the electricity price is between 0.4 and 0.6 yuan / kWh, the user will think that the electricity price is appropriate. When the electricity price is higher than 0.8 yuan / kWh, the user will think that the electricity price is expensive. Therefore, the membership function for the charging station electricity price is constructed as shown below, and the specific effect is as follows. Figure 3 As shown in Figure 2, the construction of the membership function of the vehicle SOC and the adjustment amount is similar.

[0035] The smaller the adjustment amount, the higher the response willingness of electric vehicle users, and the higher the reliability in the actual demand response process. Taking the adjustment as an example, when the response willingness is selected as 0.8 and 0.5 respectively, the results of the adjustable power lower limit are as follows: Figure 4 As shown in Figure 1. Generally, the smaller the adjustment amount, the higher the response willingness of electric vehicle users, and the higher the reliability in the actual demand response process. Select the downward adjustment response willingness as 0.8 and the upward adjustment response willingness as 0.6. When the charging electricity price is fixed within a day, the upper and lower limits of the adjustable potential of the electric vehicle cluster within a day are as follows: Figure 5 shown.

[0036] According to the different electricity consumption characteristics of flexible loads, the response willingness of aggregated users is quantified, and the membership function of flexible load influencing factors is constructed using the TSK fuzzy model. The relationship between the upward and downward response willingness of the flexible load cluster and other influencing factors is quantified. According to the determined response willingness threshold, the upper and lower limits of the adjustable potential of the flexible load cluster are calculated as one of the considerations for the reliability of the regulation quantity, providing more abundant information for the optimized operation of the distribution network, and guiding users' electricity consumption behavior in combination with the real-time electricity price plan, thereby reducing the peak load of the power grid and filling the trough load of the power grid.

[0037] Example Step 1: Conduct demand response modeling for typical flexible loads on the demand side; S11. Based on the real data of electric vehicles, analyze the adjustable ability and willingness of electric vehicles. The influencing factors include the SOC value of the electric vehicle battery, the charging fee of the charging station, and the demand-side response adjustment amount.

[0038] Step 2: Use TSK fuzzy model to establish the membership function of flexible load influencing factors; S21. Define corresponding fuzzy subsets for the three influencing factors of charging station electricity price, SOC and adjustment amount.

[0039] S22. For each group of determined input electric vehicle cluster electricity price, battery SOC and adjustment amount, the dispatch intention under a single rule is obtained. The fuzzy rule corresponding to the TSK fuzzy model is: Where: x1, x2 and x3 are the input variables charging station electricity price, electric vehicle battery SOC and regulation amount respectively; is the fuzzy subset corresponding to the input electricity price of the kth rule, is the fuzzy subset corresponding to the input variable SOC of the kth rule, is the fuzzy subset corresponding to the input variable adjustment amount of the kth rule; and is the fuzzy connection operator; f k (x) is the linear function value corresponding to the kth fuzzy rule; is the fixed parameter of the kth fuzzy rule; K is the number of fuzzy rules.

[0040] S23. For the upward and downward response willingness of electric vehicle clusters, the membership function of the fuzzy subsets corresponding to the charging station electricity price, battery SOC and adjustment amount adopts a trapezoidal distribution.

[0041] Step 3: Quantify the coupling between the upward and downward response willingness of the flexible load cluster and other influencing factors; S31. When a single influencing factor remains unchanged, the relationship between the response willingness of the electric vehicle cluster and other factors is constructed; S32. Calculate the response intention of the electric vehicle and air conditioning cluster output by the TSK fuzzy model based on the influencing factors: Where: and are the membership functions corresponding to the fuzzy subsets; EV is the response willingness value of the electric vehicle cluster.

[0042] Step 4: Set the response willingness threshold according to actual demand and calculate the upper and lower limits of the flexible load cluster's adjustable potential. Select the lower and upper response willingness thresholds and calculate the upper and lower limits of the flexible load cluster's adjustable potential.

[0043] The system of this embodiment is composed of vehicle dynamic data from Suzhou, Wuxi and Changzhou. The charging load calculation based on the Monte Carlo method calculates the charging power distribution of a certain electric vehicle cluster within a day at a 15-minute granularity. Figure 2 shown.

[0044] For the three influencing factors of charging station electricity price, SOC and regulation, the corresponding fuzzy subsets are defined respectively and the dispatch willingness under a single rule is obtained. The final output of the TSK fuzzy model is the response willingness of the electric vehicle cluster: Taking the charging station electricity price x1 as an example, when the electricity price is less than 0.2 yuan / kWh, the user will think that the electricity price is cheap, when the electricity price is between 0.4 and 0.6 yuan / kWh, the user will think that the electricity price is appropriate, and when the electricity price is higher than 0.8 yuan / kWh, the user will think that the electricity price is expensive. Therefore, the membership function for the charging station electricity price is constructed as follows, and the specific effect is as follows Figure 3 As shown in Figure 2, the construction of the membership function of the vehicle SOC and the adjustment amount is similar.

[0045] The membership function of x1 to cheap is: The appropriate membership function for x1 is: The membership function of x1 for expensive pairs is: Taking the response willingness into consideration, taking the adjustment as an example, when the response willingness is selected as 0.8 and 0.5 respectively, the results of the adjustable power lower limit are as follows: Figure 4 As shown in Figure 1. Generally, the smaller the adjustment amount, the higher the response willingness of electric vehicle users, and the higher the reliability in the actual demand response process. Select the downward adjustment response willingness as 0.8 and the upward adjustment response willingness as 0.6. When the charging electricity price is fixed within a day, the upper and lower limits of the adjustable potential of the electric vehicle cluster within a day are as follows: Figure 5 shown.

[0046] When collecting demand-side response adjustment, the following steps are mainly included: (1) Data Collection First of all, we need to collect relevant data that affect the prediction and reporting results, which can be divided into the following categories: 1) Weather data in the area where the load aggregator’s charging piles are located, including temperature, rainfall, snowfall, etc., with a refresh frequency of 1 hour. These data are used for model training and prediction.

[0047] 2) Charging pile information, including AC / DC pile attributes, bus-specific piles / public charging piles / community public charging piles / community private charging pile attributes, fixed agreement signing / flexible contract signing attributes, rated power and other attributes, corresponding platform ID, standby / shutdown / running / normal / adjustment status, and the grid node where the pile is located.

[0048] 3) Charging user information, including the vehicle’s private car / online car-hailing / bus attributes, whether the regulation is accepted, the vehicle’s SOC status, the charging target (battery percentage), and the default charging power.

[0049] 4) Market related information, including real-time electricity price information and oil price information. After the spot market is opened, it also includes 96 node electricity price information within the day and the day before.

[0050] Preprocess, store and model the collected relevant data information.

[0051] 1) Basic data processing and classification The most basic load data is one node every 15 minutes. When making hour-ahead forecasts within a day, it is necessary to average the power of the four time nodes in each hour to obtain the historical load data at 24 o'clock every day.

[0052] 2) Processing of missing data Since the vehicle-grid interaction is still in the pilot preparation stage, there is often a problem of missing data in practice. Therefore, special processing of incomplete data is required. In this solution, the data of the two closest similar time points is found for processing. Standardizing and preprocessing the data is conducive to improving the training speed and prediction accuracy of the model. Commonly used data standardization methods include maximum and minimum value standardization and mean variance standardization.

[0053] 3) Extraction and classification of relevant data Then extract indicator information that affects the prediction results, such as SOC, weather information, extreme holidays, major events, etc.

[0054] 4) Load forecasting model training classifies and groups the adjustable charging piles connected to the platform into public AC piles for industrial and commercial purposes, public DC piles, dedicated bus charging piles, community DC public piles, community AC public piles, and community private car dedicated AC piles.

[0055] For day-ahead forecasting, it is necessary to forecast the data of 96 nodes of the next day one day in advance. This is a multi-step forecasting, which is relatively difficult and requires more relevant data and a more complex forecasting model. In this technical solution, the day-ahead load forecasting adopts the WT-WPCA-LSTM hybrid forecasting model. Among them, WT is the abbreviation of Wavelet Transform, WPCA is the abbreviation of Wavelet Principle Component Analysis, and LSTM is the abbreviation of Long Short-Term Memory. The basic idea of ​​modeling is to organically combine the three models of WT denoising, WPCA wavelet principal component analysis and LSTM.

[0056] a. Wavelet transform WT The prediction of electric vehicle charging load series is easily affected by noise. Therefore, before modeling and analyzing the charging load time series data, its historical information should be denoised. The fundamental purpose of denoising is to remove noise information as much as possible while retaining the main data characteristics of the original signal. Traditional denoising methods include smoothing filtering, moving average filtering, simple nonlinear noise elimination, and linear Fourier filtering. Electric vehicle charging load time series data not only contains noise, but also generally has the characteristics of nonlinearity, non-stationarity, and high volatility. However, these traditional denoising methods cannot effectively remove the noise components in the electric vehicle charging load time series data. Wavelet transform technology has the ability to process non-stationary and irregular load time series data and retain more original information. Therefore, it is necessary to use wavelet transform technology to perform wavelet denoising on the basic historical information historical load data P that directly affects the prediction results in order to obtain more trend information.

[0057] b. Wavelet principal component analysis WPCA The WPCA denoising method is achieved by directly expanding the ordinary wavelet denoising method. First, multiple ordinary wavelets are used to perform wavelet decomposition on each dimension of the multivariate signal to obtain the decomposed wavelet coefficient matrix. Then, the overall noise size is estimated by the wavelet coefficient matrix, and a suitable threshold is selected to perform threshold processing on the wavelet coefficient matrix. Next, the principal component analysis technique is used to extract the principal components of the thresholded wavelet coefficient matrix and remove the minor components in the wavelet coefficient matrix. Finally, the wavelet coefficient matrix with the minor components removed is subjected to an inverse wavelet transform to reconstruct the original multivariate signal. In short, the WPCA denoising method uses ordinary wavelets to denoise the multivariate signal while using PCA technology to retain the main components of the original multidimensional data and remove the minor components, thereby achieving a stronger denoising effect, while also retaining the relevant information between the original multivariate signals.

[0058] c.LSTM The electric vehicle charging load time series data has long memory, while ordinary neural networks do not have memory capabilities. Traditional neural networks believe that the input samples are independent of each other, that is, the current output is only related to the current input, but not to the previous input. Therefore, traditional neural networks cannot effectively utilize sequence information, and the effect of time series analysis and processing is not ideal. Although the recursive neural network RNN ​​can model time series data well, it is difficult to learn the long-term dependencies of time series due to the problems of gradient explosion and gradient vanishing. The long short-term memory network (LSTM) is an effective structure that uses the state of memory cells to combat gradient vanishing and gradient explosion.

[0059] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0060] The specific embodiments described herein are merely examples of the spirit of the present invention. The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that technicians in the technical field to which the present invention belongs can make various modifications or supplements to the described specific embodiments or replace them in a similar manner, but will not deviate from the spirit of the present invention or exceed the scope defined in the attached claims. For ordinary technicians in this field, multiple deformations and improvements can also be made without departing from the concept of the present invention. Therefore, the scope of protection of the patent of the present invention shall be based on the attached claims.

Claims

1. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness, characterized in that: The following steps are involved: S1 analyzes the factors affecting typical flexible loads on the demand side; S2 performs demand response modeling for typical flexible loads on the demand side; S3 uses TSK fuzzy model to establish the membership function of flexible load influencing factors; S4 quantifies the coupling between the upward and downward response willingness of the flexible load cluster and other influencing factors; S5 sets the response willingness threshold and calculates the upper and lower limits of the adjustable potential of the flexible load cluster.

2. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 1, characterized in that: The influencing factors of step S1 include the SOC value of the electric vehicle battery, the charging fee of the charging station and the demand side response adjustment amount.

3. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 1 or 2, characterized in that: Based on the real data of electric vehicles, the adjustable capacity and adjustment willingness of electric vehicles are analyzed, and the demand response modeling of typical flexible loads on the demand side is carried out.

4. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 1 or 2, characterized in that: The step S3 specifically includes: S3.1 defines corresponding fuzzy subsets for the three influencing factors of charging station electricity price, SOC and regulation amount; S3.2 obtains the dispatch intention under a single rule for each group of determined input electric vehicle clusters’ electricity price, battery SOC and regulation amount.

5. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 4, characterized in that: The fuzzy rules corresponding to the TSK fuzzy model in step S3.2 are: Where: x1, x2 and x3 are the input variables charging station electricity price, electric vehicle battery SOC and regulation amount respectively; is the fuzzy subset corresponding to the input electricity price of the kth rule, is the fuzzy subset corresponding to the input variable SOC of the kth rule, is the fuzzy subset corresponding to the input variable adjustment amount of the kth rule; and is the fuzzy connection operator; f k (x) is the linear function value corresponding to the kth fuzzy rule; is the fixed parameter of the kth fuzzy rule; K is the number of fuzzy rules.

6. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 1, characterized in that: The step S4 specifically includes: S4.1 When a single influencing factor remains unchanged, the relationship between the response willingness of the electric vehicle cluster and other factors is constructed; S4.2 Based on the influencing factors, the response willingness of the electric vehicle and air-conditioning cluster that is finally output by the TSK fuzzy model is obtained.

7. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 6, characterized in that: The step S4.2 specifically includes: Where: and are the membership functions corresponding to the fuzzy subsets; EV is the response willingness value of the electric vehicle cluster.

8. The method for evaluating the adjustable potential of controllable loads considering user response willingness according to claim 1, characterized in that: The step S5 specifically includes selecting a threshold value of the downward response willingness and an upward response willingness, and calculating the upper and lower limits of the adjustable potential of the flexible load cluster.

9. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 5, characterized in that: The charging station electricity price x1 sets a judgment threshold. When the electricity price is less than 0.2 yuan / kWh, the user will think the electricity price is cheap. When the electricity price is between 0.4-0.6 yuan / kWh, the user will think the electricity price is appropriate. When the electricity price is higher than 0.8 yuan / kWh, the user will think the electricity price is expensive.

10. A method for evaluating the adjustable potential of controllable loads taking into account user response willingness according to claim 8, characterized in that: The smaller the adjustment amount is, the higher the response willingness of the electric vehicle user is, and the higher the reliability is in the actual demand response process.

Images