Cogeneration user group energy efficiency optimization method and system based on bidirectional cooperation

Abstract

The invention discloses a heat and power cogeneration user group energy efficiency optimization method and system based on bidirectional cooperation, relates to the technical field of heat and power cogeneration, and provides the following scheme that the method comprises the steps of setting a historical period, collecting historical electricity and heat consumption data of a user in the period, and constructing a strategy feature vector; the strategy feature vector comprises an electricity and heat utilization behavior feature, a demand response participation degree feature and an economic benefit feature; classifying the user strategy feature vectors by adopting a clustering algorithm, and identifying information island users and main group users; the method is based on historical electricity and heat consumption data of island users. Three-dimensional features of power utilization behaviors, demand response participation degree and economic benefits are extracted through historical power utilization and heat utilization data, strategy feature vectors are constructed, accurate description of the power utilization and heat utilization behaviors of users is achieved, limitation of traditional single-dimensional analysis is broken through, an accurate data basis is provided for user clustering and strategy optimization, feature recognition accuracy is improved, and user experience is improved. And the system can accurately analyze the electricity and heat utilization rule.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of combined heat and power, in particular to a combined heat and power user group energy efficiency optimization method and system based on bidirectional collaboration. BACKGROUND

[0002] In the technical field of combined heat and power, especially in the scenario of user group energy efficiency optimization, the existing technology is based on user historical electricity and heat data to develop strategy making work, and its core idea is to analyze basic data such as power load curve and heat demand trend, extract simple statistical characteristics such as average, peak and valley of electricity load, and develop general energy distribution and dispatching strategy for the main group in combination with electricity price policy.

[0003] However, the existing technology still has significant defects: on the one hand, the description of user characteristics is limited to a single dimension, and the deep law of user behavior and benefit correlation cannot be comprehensively considered, resulting in too rough user classification and difficulty in accurately matching optimization strategies; on the other hand, the division of user groups mostly uses clustering methods with fixed parameters, which cannot effectively identify island users with special behavior patterns, and the strategy optimization is mostly one-way guidance, ignoring the potential value of high-efficiency strategies in island users, which easily makes the main group strategy fall into local optimum, therefore we propose a combined heat and power user group energy efficiency optimization method and system based on bidirectional collaboration to solve this problem. SUMMARY

[0004] To solve the above technical problems, the present application provides a combined heat and power user group energy efficiency optimization method and system based on bidirectional collaboration, which solves the problem that the existing technology still has significant defects: on the one hand, the description of user characteristics is limited to a single dimension, and the deep law of user behavior and benefit correlation cannot be comprehensively considered, resulting in too rough user classification and difficulty in accurately matching optimization strategies; on the other hand, the division of user groups mostly uses clustering methods with fixed parameters, which cannot effectively identify island users with special behavior patterns, and the strategy optimization is mostly one-way guidance, ignoring the potential value of high-efficiency strategies in island users, which easily makes the main group strategy fall into local optimum.

[0005] To achieve the above purpose, the technical scheme adopted by the present application is as follows: the present application provides a combined heat and power user group energy efficiency optimization method based on bidirectional collaboration, which comprises: Setting a historical period, collecting historical electricity and heat data of users in the period, constructing a strategy feature vector, the strategy feature vector including electricity and heat behavior characteristics, demand response participation degree characteristics and economic benefit characteristics; Classifying user strategy feature vectors using clustering algorithm to identify information island users and main group users; Based on the historical electricity and heat data of the island user, the monthly load variance and the abnormal frequency of price response are calculated, and the historical strategy path of the similar main group user is matched; The island user is guided to converge to the main group strategy path, and the strategy with the efficiency higher than the preset potential value in the island user is used to correct the historical strategy path library of the main group in reverse; A dynamic feedback mechanism is established to monitor the implementation effect of the corrected main group strategy path in real time, calculate the comprehensive efficiency index of the corrected path, and optimize the reference group path when the index is lower than the preset reference value for three consecutive periods.

[0006] Further, a heat and power cogeneration user group energy efficiency optimization system based on bidirectional collaboration is proposed for implementing any of the above-mentioned heat and power cogeneration user group energy efficiency optimization methods based on bidirectional collaboration, comprising: A strategy feature vector construction module is used to set a historical period, collect historical electricity and heat data of the user in the period, and construct a strategy feature vector, wherein the strategy feature vector includes electricity and heat behavior characteristics, demand response participation characteristics and economic benefit characteristics; A user group classification module is used to classify the user strategy feature vector by using a clustering algorithm to identify information island users and main group users; A strategy path matching module is used to calculate the monthly load variance and the abnormal frequency of price response based on the historical electricity and heat data of the island user, and match the historical strategy path of the similar main group user; A bidirectional collaborative optimization module is used to guide the island user to converge to the main group strategy path, and to correct the historical strategy path library of the main group in reverse by using the strategy with the efficiency higher than the preset potential value in the island user; A dynamic feedback and iteration module is used to establish a dynamic feedback mechanism to monitor the implementation effect of the corrected main group strategy path in real time, calculate the comprehensive efficiency index of the corrected path, and optimize the reference group path when the index is lower than the preset reference value for three consecutive periods.

[0007] Compared with the prior art, the beneficial effects of the present application are: 1. The three-dimensional characteristics of electricity consumption behavior, demand response participation and economic benefit are extracted from the historical electricity and heat data to construct a strategy feature vector, accurately describe the electricity and heat behavior of the user, break through the limitations of traditional single-dimensional analysis, provide accurate data basis for user clustering and strategy optimization, improve the accuracy of feature recognition, and help the system accurately analyze the electricity and heat rules.

[0008] 2. By matching similar main group path, guiding the island user to converge to the efficient strategy, and at the same time, correcting the main group path library by the strategy with efficiency higher than the preset potential value in the island user, the low efficiency behavior of the island user is eliminated, the high quality strategy of the small group is mined, the local optimal path of the main group is avoided, and the comprehensiveness and efficiency of the system strategy are improved.

[0009] 3. Based on the reference group path, the load adjustment range and the electricity price trigger threshold are extracted, the adjustment suggestion is pushed when the real-time electricity price is higher than the threshold, the economic benefit is calculated and optimized by combining the time sequence electricity cost, the response timeliness and accuracy of the user to the electricity price signal are improved, and the system operation pressure is reduced and the user electricity cost is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 A flow chart of a heat and power cogeneration user group energy efficiency optimization method based on bidirectional cooperation is provided for the present application; Figure 2 A method flow chart for constructing a strategy feature vector in the present application; Figure 3 A method flow chart for matching historical strategy paths in the present application; Figure 4 A structure block diagram of a heat and power cogeneration user group energy efficiency optimization system based on bidirectional cooperation is provided for the present application. DETAILED DESCRIPTION

[0011] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be thought of by those skilled in the art.

[0012] Referring to Figures 1-3 The heat and power cogeneration user group energy efficiency optimization method based on bidirectional cooperation, the method comprises: Set the historical period, collect the historical electricity and heat data of the user in the period, construct the strategy feature vector, and the strategy feature vector contains the electricity and heat behavior characteristics, the demand response participation characteristics and the economic benefit characteristics; Classify the user strategy feature vector by using clustering algorithm, and identify the information island user and the main group user; Based on the historical electricity and heat data of the island user, the monthly load variance and the electricity price response abnormal frequency are calculated, and the historical strategy path of the similar main group user is matched; Guide the island user to converge to the main group strategy path, and at the same time, correct the historical strategy path library of the main group by the strategy with efficiency higher than the preset potential value in the island user; A dynamic feedback mechanism is established to monitor the implementation effect of the revised strategy path of the main group in real time, calculate the comprehensive performance index of the revised path, and optimize the reference group path when the index is lower than the preset reference value for three consecutive periods.

[0013] Referring to Figure 2 As shown in the figure, the constructed strategy feature vector specifically includes: Based on historical electricity and heat consumption data, the power load curve and heat demand data curve are drafted, and the corresponding statistical indicators and time periodicity characteristics are extracted; The statistical indicators include mean, variance, peak load and valley value proportion, and the time periodicity characteristics include frequency components of main cycle amplitude and their phases, and the electricity and heat consumption behavior characteristics are jointly represented by statistical indicators and time series periodicity characteristics; Exemplarily, the historical period is set to one year, the power load curve is drawn by a sampling frequency of 15 minutes / point, the daily average load is calculated as the mean, the daily load variance is counted, the monthly maximum load and the proportion of the period within a year when the load is lower than the mean are recorded; the load curve is subjected to Fourier transform, the main frequency components within a day are extracted, and the peak value occurrence time is determined through phase analysis, such as daily 14:00 as the starting phase of the electricity peak, and the reference electricity and heat consumption behavior characteristics can be expressed as [mean=12.5, variance=8.2, peak=28, valley value proportion=0.35, main frequency amplitude=0.6, phase=14]; The participation rate, average response time and average response amount of the user participating in the price response project within the historical period are counted, wherein the participation rate is the ratio of the response times to the price adjustment times; The demand response participation degree feature is jointly represented by the participation rate, average response time and average response amount; Exemplarily, the historical period is one year, i.e. the price is adjusted 30 times within a year, the response times of the user is 15 times, and the participation rate is 50%. The average delay time from the price release to the load adjustment and the average load change amount of each response are recorded, and the reference demand response participation degree feature can be expressed as [participation rate=0.5, average response time=2.3, average response amount=4.5]; Based on the real-time price and the subsidy of the response project, the actual price of different time periods is obtained, the time series electricity cost of the user within the historical period is obtained by combining the historical electricity and heat consumption data, and is taken as the economic benefit feature; Exemplarily, the real-time price is obtained, such as 1.2 yuan / kwh during peak time and 0.5 yuan / kwh during valley time, the subsidy of the response project is 0.05 yuan / kwh, the proportion of each time period in the historical electricity and heat consumption data is obtained, and the electricity cost of each time period in the day is calculated, the monthly electricity cost is accumulated, and the time series electricity cost is arranged in the order of months.

[0014] Specifically, the clustering algorithm is used to classify the user strategy feature vector, and identify information island users and main group users, specifically comprising: According to the user group characteristics of the combined heat and power system, the value range of the cluster number is set, and for each cluster number value, the K-means++ algorithm is used to select the initial cluster center; For example, in the actual scene, users can be divided into 3-5 typical groups, such as high response users, low response users, seasonal fluctuation users, information island users, etc., and the cluster number value range can be appropriately extended to 2-6 on this basis; Calculate the Euclidean distance of the strategy feature vector of each user to the cluster center, and assign it to the nearest cluster, and update the cluster center until convergence; Calculate the silhouette coefficient of the user in the clustering result, draw the average silhouette coefficient curve corresponding to different cluster number values, and select the cluster number value that maximizes the silhouette coefficient as the final clustering cluster number; If the average silhouette coefficient of the users in a cluster is lower than a set threshold, or the proportion of the number of users in the cluster is less than a set proportion, the cluster is determined to be an information island user group, and if not, the cluster is determined to be a main group user group; For example, the threshold can be 0.2, and the set proportion can be 5%; The calculation formula of the silhouette coefficient is: ; In the formula, is the silhouette coefficient, is the average Euclidean distance of user i and other users in the same cluster, is the minimum average Euclidean distance of user i to other clusters; It should be noted that the user strategy features in the combined heat and power system are complex, and if a fixed cluster number is directly used, some island users may be misclassified as normal user clusters, or normal users may be misjudged as island users due to feature fluctuations; Using the cluster number that maximizes the silhouette coefficient can accurately capture the real island user group and avoid over-segmentation.

[0015] Referring to Figure 3 Based on the historical electricity and heat data of the island users, the monthly load variance and the abnormal frequency of price response are calculated, and the historical strategy path of similar main group users is matched, specifically comprising: Extract the historical electricity and heat data of the island users, calculate the monthly load variance, and set a delay threshold, and calculate the abnormal frequency of price response combined with the demand response participation characteristics; Vectorize the monthly load variance and the abnormal frequency of price response of the main group users to form a historical strategy path library, and each path contains seasonal fluctuation patterns and price response behaviors; It should be noted that the seasonal fluctuation mode is the law that the user's electricity demand changes periodically with the factors of season, climate and holidays in the historical period, and the quantitative indicators include monthly load variance, daily peak-valley difference rate and seasonal index, the seasonal index is extracted by decomposing the historical electricity and heat data by time series; the price response behavior is the tendency of the user to adjust the energy strategy according to the price signal, and the quantitative indicators are the price elasticity coefficient, the price response abnormal frequency and the delay time, the price elasticity coefficient is the ratio of electricity consumption to price, and the delay time is the difference between the user strategy adjustment time and the price release time; The dynamic time warping algorithm is used to calculate the similarity of the characteristic sequences of the island user and the main group user, and the main group user with the highest similarity is selected as the reference group; Based on the historical strategy path of the reference group, the load adjustment amplitude and the price adjustment trigger threshold are extracted, and the strategy adjustment baseline of the island user is generated accordingly; It should be noted that the load adjustment amplitude is the load change range of the island user, and the price adjustment trigger threshold is the response critical value of the user to the price change, when the real-time price fluctuation exceeds the threshold, the user needs to trigger the strategy adjustment; It is further explained that the monthly load variance is used to lock the long-term mode of the user, and the price response abnormal frequency is used to verify the consistency of the user strategy, and the combination of the two can efficiently select the main user group similar to the behavior of the island user and reliable in strategy.

[0016] Specifically, the island user is guided to converge to the main group strategy path, and the strategy with higher efficiency than the preset potential value in the island user is used to correct the historical strategy path library of the main group in reverse, specifically including: The current electricity and heat data and the price response project data of the island user are collected in real time, and the real-time price is extracted from the price response project data; If the real-time price exceeds the price adjustment trigger threshold in the strategy adjustment baseline, the load adjustment amplitude is pushed to the island user; After the user adjusts the electricity consumption behavior according to the load adjustment amplitude, the ratio of the monthly load data to the monthly electricity cost is taken as the user efficiency; The strategy adjustment baseline with the user efficiency greater than the preset potential value is weighted and combined with the strategy path of the corresponding main group to obtain the corrected strategy path, and the corrected strategy path is included in the historical strategy path library of the main group user; for example, the preset potential value can be 10%.

[0017] Specifically, the dynamic feedback mechanism is established to monitor the implementation effect of the corrected main group strategy path in real time, calculate the comprehensive efficiency index of the corrected path, and optimize the reference group path when the index is lower than the preset reference value for three consecutive periods, specifically including: Set a feedback period, collect the actual electricity and heat consumption data and strategy execution records of the main group users in each feedback period, and calculate the average efficiency improvement rate of the users; Specifically, the average efficiency improvement rate = (user efficiency after strategy execution - user efficiency before strategy execution) / user efficiency before strategy execution; Calculate the proportion of the number of users executing the strategy in the total number of main group users as the proportion of strategy-covered users; The product of the average efficiency improvement rate and the proportion of strategy-covered users is the comprehensive efficiency index. When the comprehensive efficiency index is lower than the preset reference value for three consecutive periods, the dynamic time warping algorithm is called again to match a more optimal reference group path. For example, the preset reference value is set according to 80% of the historical optimal comprehensive efficiency index. For example, the way to rematch the more optimal reference group path is: select the top three main group users in similarity as the reference group, use the electricity consumption behaviors of these three users in the last feedback period as the reference vector, calculate the similarity of the feature sequence of each path in the strategy path library of the three users through the dynamic time warping algorithm, select the top three paths with the highest similarity as candidate paths, and select the path with the highest comprehensive efficiency index as the more optimal reference group path.

[0018] Referring to Figure 4 The scheme proposes a heat and power cogeneration user group energy efficiency optimization system based on bidirectional collaboration, which is used to implement the above-mentioned heat and power cogeneration user group energy efficiency optimization method based on bidirectional collaboration, and includes: A strategy feature vector construction module is configured to set a historical period, collect historical electricity and heat consumption data of users in the period, and construct a strategy feature vector. The strategy feature vector includes electricity and heat consumption behavior features, demand response participation features, and economic benefit features. A user group classification module is configured to classify user strategy feature vectors using a clustering algorithm to identify information island users and main group users. A strategy path matching module is configured to calculate monthly load variance and price response abnormal frequency based on historical electricity and heat consumption data of island users, and match historical strategy paths of similar main group users. A bidirectional collaborative optimization module is configured to guide island users to converge to the main group strategy path, and simultaneously use strategies with efficiency higher than a preset potential value in the island users to correct the historical strategy path library of the main group. A dynamic feedback and iteration module is configured to establish a dynamic feedback mechanism, monitor the implementation effect of the main group strategy path after correction in real time, calculate the comprehensive efficiency index of the corrected path, and optimize the reference group path when the index is lower than a preset reference value for three consecutive periods.

[0019] The strategy feature vector construction module specifically includes: A multi-dimensional feature extraction unit is used to formulate power load curves and heat demand data curves based on historical electricity and heat consumption data, and to extract corresponding statistical indicators and time periodic features; The demand response characteristic statistics unit is used to count the participation rate, average response time and average response volume of users in the electricity price response project during the historical period. The participation rate is the ratio of the number of responses to the number of electricity price adjustments. The economic benefit calculation unit is used to obtain the actual electricity prices in different time periods based on the real-time electricity prices and the subsidies of the response projects, and combine them with the historical electricity and heat consumption data to obtain the user's time-series electricity consumption costs in the historical period.

[0020] The user group classification module specifically includes: The clustering parameter adaptive unit is used to set the value range of the number of clusters according to the user group characteristics of the cogeneration system, and for each cluster number value, the K-means++ algorithm is used to select the initial cluster center; The silhouette coefficient analysis unit is used to calculate the Euclidean distance between each user's strategy feature vector and the center of each cluster, assign it to the cluster with the closest distance, and update the cluster center until convergence.

[0021] The dynamic feedback and iteration module specifically includes: The efficiency index calculation subunit is used to set the feedback cycle. During each feedback cycle, the actual electricity and heat consumption data and policy execution records of the main group of users are collected, and the average efficiency improvement rate of users is calculated. The proportion of users who implement the policy to the total number of users in the main group is counted as the proportion of users covered by the policy. The optimal path rematching unit is used to multiply the average performance improvement rate and the proportion of users covered by the strategy as the comprehensive performance index. When the comprehensive performance index is lower than the preset benchmark value for three consecutive cycles, the dynamic time warping algorithm is re-called to match a better reference group path.

[0022] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing energy efficiency of combined heat and power user groups based on bidirectional collaboration, characterized in that: The method comprises: Set a historical period, collect historical electricity and heat consumption data of users within the period, and construct a strategy feature vector. The strategy feature vector includes electricity and heat consumption behavior characteristics, demand response participation characteristics, and economic benefit characteristics. Clustering algorithms are used to classify user strategy feature vectors and identify information island users and main group users; Based on the historical electricity and heat consumption data of isolated users, the monthly load variance and abnormal frequency of electricity price response are calculated, and the historical strategy paths of similar main group users are matched; Guide the isolated users to converge to the main group's strategy path, and at the same time use the strategies of the isolated users with higher effectiveness than the preset potential value to reversely correct the historical strategy path library of the main group; Establish a dynamic feedback mechanism to monitor the implementation effect of the revised main group strategy path in real time, calculate the comprehensive effectiveness index of the revised path, and optimize the reference group path when the index is lower than the preset benchmark value for three consecutive cycles.

2. The method according to claim 1, characterized in that The construction strategy feature vector specifically includes: Based on historical electricity and heat consumption data, the power load curve and heat demand data curve are formulated, and the corresponding statistical indicators and time periodic characteristics are extracted; The statistical indicators include mean, variance, peak load and valley ratio; the time periodic characteristics include the frequency component of the main cycle amplitude and its phase; the electricity and heat consumption behavior characteristics are jointly characterized by statistical indicators and time series periodic characteristics; Collect statistics on the participation rate, average response time, and average response volume of users in the electricity price response project during the historical period. The participation rate is the ratio of the number of responses to the number of electricity price adjustments. The demand response participation characteristic is jointly characterized by participation rate, average response time and average response volume; Based on the real-time electricity price and the subsidy of the response project, the actual electricity price in different time periods is obtained. Combined with the historical electricity and heat consumption data, the user's time-series electricity cost in the historical period is obtained and used as the economic benefit feature.

3. The method according to claim 2, characterized in that The clustering algorithm is used to classify the user strategy feature vectors and identify information island users and main group users, specifically including: According to the user group characteristics of the cogeneration system, the value range of the number of clusters is set, and for each cluster value, the K-means++ algorithm is used to select the initial cluster center; Calculate the Euclidean distance between each user's strategy feature vector and the center of each cluster, assign it to the cluster with the closest distance, and update the cluster center until convergence; Calculate the silhouette coefficient of the user in the clustering results, draw the average silhouette coefficient curve corresponding to different cluster number values, and select the cluster number value that maximizes the silhouette coefficient as the final cluster number; If the average silhouette coefficient of a cluster of users is lower than the set threshold, or the number of users in the cluster is less than the set ratio, the cluster is determined to be an information island user group. If neither is the case, the cluster is determined to be a main group user group. The calculation formula of the silhouette coefficient is: ； Where, is the silhouette coefficient, is the average Euclidean distance between user i and other users in the same cluster, is the minimum average Euclidean distance from user i to other clusters.

4. The method according to claim 1, wherein The calculation of monthly load variance and abnormal frequency of electricity price response based on the historical electricity and heat consumption data of isolated users, and matching the historical strategy paths of similar main group users, specifically includes: Extract historical electricity and heat consumption data of isolated users, calculate their monthly load variance, set a delay threshold, and calculate the abnormal frequency of their electricity price response based on the demand response participation characteristics; The monthly load variance and price response anomaly frequency of the main group of users are vectorized to form a historical strategy path library, each of which contains seasonal fluctuation patterns and price response behaviors; The dynamic time warping algorithm is used to calculate the similarity of the feature sequences of isolated users and main group users, and the main group users with the highest similarity are selected as the reference group; Based on the historical strategy path of the reference group, its load adjustment range and electricity price adjustment trigger threshold are extracted to generate the strategy adjustment baseline for the isolated users.

5. The method according to claim 4, characterized in that The method of guiding the isolated users to converge to the main group's strategy path and simultaneously using the strategies of the isolated users with higher effectiveness than the preset potential value to reversely correct the main group's historical strategy path library specifically includes: Collect the current electricity and heat consumption data and electricity price response project data of isolated users in real time, and extract the real-time electricity price from the electricity price response project data; If the real-time electricity price exceeds the electricity price adjustment trigger threshold in the strategy adjustment baseline, the load adjustment range will be pushed to the isolated island user; After the user adjusts his electricity consumption behavior according to the load adjustment range, the ratio of his monthly load data to the monthly electricity cost is used as the user effectiveness; The policy adjustment baseline of users whose effectiveness is greater than the preset potential value is weightedly merged with the corresponding main group's policy path to obtain a revised policy path, which is then included in the historical policy path library of the main group users.

6. The method according to claim 5, characterized in that The dynamic feedback mechanism is established to monitor the implementation effect of the revised main group strategy path in real time and calculate the comprehensive effectiveness index of the revised path. When the index is lower than the preset benchmark value for three consecutive cycles, the reference group path is optimized, specifically including: Set a feedback cycle. During each feedback cycle, collect the actual electricity and heat consumption data and policy execution records of the main group of users, and calculate the average efficiency improvement rate of users. Count the proportion of users who have executed the policy to the total number of users in the main group, and use this as the proportion of users covered by the policy. The product of the average performance improvement rate and the proportion of users covered by the strategy is the comprehensive performance index. When the comprehensive performance index is lower than the preset benchmark value for three consecutive cycles, the dynamic time warping algorithm is re-called to match a better reference group path.

7. A cogeneration user group energy efficiency optimization system based on two-way collaboration, characterized in that: The method for optimizing energy efficiency of a combined heat and power user group based on bidirectional collaboration according to any one of claims 1 to 6 comprises: A strategy feature vector construction module is used to set a historical period, collect historical electricity and heat consumption data of users during the period, and construct a strategy feature vector. The strategy feature vector includes electricity and heat consumption behavior characteristics, demand response participation characteristics, and economic benefit characteristics. User group classification module, which is used to classify user strategy feature vectors using clustering algorithms to identify information island users and main group users; The strategy path matching module is used to calculate the monthly load variance and abnormal frequency of electricity price response based on the historical electricity and heat consumption data of isolated users, and match the historical strategy paths of similar main group users; A bidirectional collaborative optimization module is used to guide isolated users to converge to the main group's strategy path. At the same time, it uses the strategies of isolated users with higher effectiveness than the preset potential value to reversely correct the main group's historical strategy path library. The dynamic feedback and iteration module is used to establish a dynamic feedback mechanism, monitor the implementation effect of the revised main group strategy path in real time, calculate the comprehensive effectiveness index of the revised path, and optimize the reference group path when the index is lower than the preset benchmark value for three consecutive cycles.

8. The energy efficiency optimization system for combined heat and power user groups based on bidirectional collaboration according to claim 7 is characterized in that: The strategy feature vector construction module specifically includes: A multi-dimensional feature extraction unit is used to formulate power load curves and heat demand data curves based on historical electricity and heat consumption data, and to extract corresponding statistical indicators and time periodic features; The demand response characteristic statistics unit is used to count the participation rate, average response time and average response volume of users in the electricity price response project during the historical period. The participation rate is the ratio of the number of responses to the number of electricity price adjustments. The economic benefit calculation unit is used to obtain the actual electricity prices in different time periods based on the real-time electricity prices and the subsidies of the response projects, and combine them with the historical electricity and heat consumption data to obtain the user's time-series electricity consumption costs in the historical period.

9. The energy efficiency optimization system for combined heat and power user groups based on bidirectional collaboration according to claim 7 is characterized in that: The user group classification module specifically includes: The clustering parameter adaptive unit is used to set the value range of the number of clusters according to the user group characteristics of the cogeneration system, and for each cluster number value, the K-means++ algorithm is used to select the initial cluster center; The silhouette coefficient analysis unit is used to calculate the Euclidean distance between each user's strategy feature vector and the center of each cluster, assign it to the cluster with the closest distance, and update the cluster center until convergence.

10. The energy efficiency optimization system for combined heat and power user groups based on bidirectional collaboration according to claim 7, characterized in that: The dynamic feedback and iteration module specifically includes: The efficiency index calculation subunit is used to set the feedback cycle. During each feedback cycle, the actual electricity and heat consumption data and policy execution records of the main group of users are collected, and the average efficiency improvement rate of users is calculated. The proportion of users who implement the policy to the total number of users in the main group is counted as the proportion of users covered by the policy. The optimal path rematching unit is used to multiply the average performance improvement rate and the proportion of users covered by the strategy as the comprehensive performance index. When the comprehensive performance index is lower than the preset benchmark value for three consecutive cycles, the dynamic time warping algorithm is re-called to match a better reference group path.

Images