A hybrid energy storage system layered joint scheduling method, system and medium

Abstract

The application provides a kind of mixed energy storage system hierarchical joint scheduling method, system and medium, it is related to energy storage system technical field, wherein the method comprises: by the output data of target mixed energy storage system in historical time is collected in two channels, obtain electrochemical energy storage output data sequence and gravity energy storage output data sequence, carry out output trend prediction, obtain electrochemical energy storage output trend characteristics and gravity energy storage output trend characteristics, carry out operating state monitoring to target mixed energy storage system, obtain real-time system operating state monitoring data, carry out multi-objective hierarchical joint scheduling optimization in combination with electrochemical energy storage output trend characteristics and gravity energy storage output trend characteristics, obtain joint scheduling optimization scheme.The technical problems that the prior art exists energy storage scheduling strategy single, resulting in the real-time and adaptability of scheduling scheme is poor are solved.The technical effects of improving the real-time and adaptability of scheduling scheme are achieved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of energy storage systems, in particular to a hierarchical joint scheduling method and system for a hybrid energy storage system and a medium. BACKGROUND

[0002] With large-scale grid connection of new energy, the fluctuation and intermittency characteristics of the power network are gradually enhanced, which brings great challenges to the safety and stability of power operation. As one of the key technologies for solving the problem of grid connection of new energy, the energy storage system is widely deployed in new energy grid connection scenarios, such as wind power plants or photovoltaic power plants, and can also be directly connected to grid scheduling nodes to effectively smooth the fluctuations of new energy generation and improve the acceptance capacity and stability of the power grid.

[0003] At present, common energy storage systems include electrochemical energy storage and gravity energy storage. Electrochemical energy storage has the advantages of fast response speed and flexible adjustment, but has defects such as limited capacity and service life, and high cost. Gravity energy storage has the characteristics of large energy storage capacity, low cost and long service life, but its dynamic response speed is relatively slow. The existing hybrid energy storage method mainly adopts fixed proportion allocation or simple threshold switching, which has the problems of single scheduling strategy and failure to fully utilize the complementary advantages of fast response of electrochemical energy storage and large capacity of gravity energy storage, resulting in poor real-time performance and adaptability of the scheduling scheme and affecting the efficiency and reliability of the power grid operation.

[0004] In summary, the prior art has the technical problem of single energy storage scheduling strategy, which leads to poor real-time performance and adaptability of the scheduling scheme. SUMMARY

[0005] The purpose of the present application is to provide a hierarchical joint scheduling method and system for a hybrid energy storage system and a medium, which solves the technical problem of single energy storage scheduling strategy in the prior art, which leads to poor real-time performance and adaptability of the scheduling scheme.

[0006] In view of the above problems, the present application provides a hierarchical joint scheduling method and system for a hybrid energy storage system and a medium.

[0007] In a first aspect, the present application provides a hierarchical joint scheduling method for a hybrid energy storage system, which comprises: collecting output data of a target hybrid energy storage system in a historical time through double channels to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence; performing output trend prediction by traversing the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to obtain electrochemical energy storage output trend characteristics and gravity energy storage output trend characteristics; monitoring the running state of the target hybrid energy storage system to obtain real-time system running state monitoring data, and combining the electrochemical energy storage output trend characteristics and the gravity energy storage output trend characteristics to perform multi-objective hierarchical joint scheduling optimization to obtain a joint scheduling optimization scheme.

[0008] Optionally, a set of historical energy storage abnormal data of the target hybrid energy storage system is obtained, a first acquisition bandwidth and a second acquisition bandwidth are determined based on the set of historical energy storage abnormal data, a first acquisition channel and a second acquisition channel are constructed based on the first acquisition bandwidth and the second acquisition bandwidth respectively, a double-channel acquisition is obtained by connecting the first acquisition channel and the second acquisition channel in parallel and connecting an output layer to a full connection layer respectively, and output data of the target hybrid energy storage system in a historical time is acquired by using the double-channel acquisition to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence.

[0009] Optionally, an interval time set is obtained by traversing and extracting abnormal interval times in the set of historical energy storage abnormal data, a maximum value in the interval time set is taken as the first acquisition bandwidth, and a minimum value in the interval time set is taken as the second acquisition bandwidth.

[0010] Optionally, a multi-scale energy storage predictor is used to respectively perform output trend prediction on the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to obtain a multi-scale electrochemical energy storage output trend feature set and a multi-scale gravity energy storage output trend feature set, the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are respectively subjected to leading interactive enhancement to determine a multi-scale enhanced electrochemical energy storage output trend feature set and a multi-scale enhanced gravity energy storage output trend feature set, and a mean value of the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set is calculated to obtain the electrochemical energy storage output trend feature and the gravity energy storage output trend feature.

[0011] Optionally, leading multi-scale electrochemical energy storage output trend features and leading multi-scale gravity energy storage output trend features are extracted from the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set, the leading multi-scale electrochemical energy storage output trend features and the leading multi-scale gravity energy storage output trend features are respectively features with the highest overall similarity to other multi-scale electrochemical energy storage output trend features and multi-scale gravity energy storage output trend features in the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set, and the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are subjected to leading interactive enhancement based on the leading multi-scale electrochemical energy storage output trend features and the leading multi-scale gravity energy storage output trend features to determine the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set.

[0012] Optionally, a fine-grained similarity degree group set is obtained by calculating the fine-grained similarity degree of the leading multi-scale electrochemical energy storage output trend feature to each multi-scale electrochemical energy storage output trend feature in the multi-scale electrochemical energy storage output trend feature set; a set of electrochemical leading interaction enhancement matrices is determined based on the fine-grained similarity degree group set; the multi-scale electrochemical energy storage output trend feature set is enhanced by using the set of electrochemical leading interaction enhancement matrices, to obtain a multi-scale enhanced electrochemical energy storage output trend feature set; and the multi-scale enhanced gravity energy storage output trend feature set is determined by leading and interaction enhancing the multi-scale gravity energy storage output trend feature set based on the leading multi-scale gravity energy storage output trend feature.

[0013] Optionally, a joint scheduling optimization layer is used to perform hierarchical optimization based on the real-time system operation state monitoring data, the electrochemical energy storage output trend feature, and the gravity energy storage output trend feature, to obtain a joint scheduling optimization scheme, with a scheduling optimization multi-objective as a constraint; the joint scheduling optimization layer includes a gravity energy storage scheduling layer at an hourly level and an electrochemical energy storage scheduling layer at a minute level.

[0014] Optionally, the scheduling optimization multi-objective is to reduce system operation cost and improve gravity energy storage consumption rate.

[0015] In a second aspect of the present application, a hybrid energy storage system hierarchical joint scheduling system is provided, which includes: a data acquisition module configured to perform double-channel acquisition on output data of a target hybrid energy storage system in a historical time, to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence; a trend prediction module configured to perform output trend prediction on the electrochemical energy storage output data sequence and the gravity energy storage output data sequence, to obtain an electrochemical energy storage output trend feature and a gravity energy storage output trend feature; and a scheduling optimization module configured to perform operation state monitoring on the target hybrid energy storage system, to obtain real-time system operation state monitoring data, and to perform multi-objective hierarchical joint scheduling optimization in combination with the electrochemical energy storage output trend feature and the gravity energy storage output trend feature, to obtain a joint scheduling optimization scheme.

[0016] In a third aspect of the present application, a computer readable storage medium is provided, which stores a computer program, and the computer program realizes the steps of the above-described hybrid energy storage system hierarchical joint scheduling method when executed.

[0017] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0018] The method provided by the embodiment of the application obtains the electrochemical energy storage output data sequence and the gravity energy storage output data sequence by collecting the output data of the target hybrid energy storage system in a historical time; the electrochemical energy storage output trend feature and the gravity energy storage output trend feature are obtained by traversing the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to perform output trend prediction; the real-time system operation state monitoring data is obtained by monitoring the operation state of the target hybrid energy storage system, and the electrochemical energy storage output trend feature and the gravity energy storage output trend feature are combined to perform multi-target hierarchical joint scheduling optimization to obtain a joint scheduling optimization scheme. The technical effects of effectively improving the real-time performance and adaptability of the scheduling scheme and enhancing the stability and reliability of power grid operation are achieved by using the hierarchical joint scheduling method, combining the rapid response of the electrochemical energy storage and the large-capacity characteristics of the gravity energy storage, and realizing multi-target optimization scheduling.

[0019] The above description is only a summary of the technical solutions of the application. In order to more clearly understand the technical means of the application, the specific embodiments of the application can be implemented according to the content of the description, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the application, nor is it intended to limit the scope of the application. Other features of the application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and other drawings can be obtained by the provided drawings without creative labor for those skilled in the art.

[0021] Figure 1 A flowchart of a hierarchical joint scheduling method of a hybrid energy storage system provided by the application.

[0022] Figure 2 A structure diagram of a hierarchical joint scheduling system of a hybrid energy storage system provided by the application.

[0023] Explanation of reference signs: data acquisition module 11, trend prediction module 12, scheduling optimization module 13. DETAILED DESCRIPTION

[0024] This application provides a hierarchical joint scheduling method, system, and medium for hybrid energy storage systems, addressing the technical problem that existing technologies suffer from poor real-time performance and adaptability due to the reliance on a single energy storage scheduling strategy. By combining the rapid response of electrochemical energy storage with the large capacity of gravity energy storage through a hierarchical joint scheduling method, multi-objective optimized scheduling is achieved, effectively improving the real-time performance and adaptability of the scheduling scheme, and enhancing the stability and reliability of power grid operation.

[0025] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. It should also be noted that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, not all of them.

[0026] Example 1, as Figure 1 As shown, this application provides a hierarchical joint scheduling method for a hybrid energy storage system, which includes:

[0027] The output data of the target hybrid energy storage system over a historical period were acquired through dual-channel acquisition to obtain the output data sequences of electrochemical energy storage and gravity energy storage.

[0028] Furthermore, a historical energy storage anomaly data set of the target hybrid energy storage system is obtained, and a first acquisition bandwidth and a second acquisition bandwidth are determined based on the historical energy storage anomaly data set; a first acquisition channel and a second acquisition channel are constructed based on the first acquisition bandwidth and the second acquisition bandwidth, respectively; by connecting the first acquisition channel and the second acquisition channel in parallel and connecting the output layer to the fully connected layer, a dual acquisition channel is obtained; the dual acquisition channel is used to acquire the output data of the target hybrid energy storage system over a historical period to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence.

[0029] Specifically, the target hybrid energy storage system includes an electrochemical energy storage unit and a gravity energy storage unit. The electrochemical energy storage unit is based on electrochemical principles and converts chemical energy and electrical energy through chemical reactions to store energy. Depending on the material, it can be in the form of lithium-ion batteries, lead-acid batteries, and flow batteries. The gravity energy storage unit stores energy by lifting or releasing heavy objects. It uses mechanical energy storage technology to lift heavy objects to high places to store potential energy. During peak electricity demand, the generator releases the heavy objects to generate electricity. The energy storage medium includes solid substances and water, which converts electrical energy and gravitational potential energy through vertical height differences. Through the monitoring system of the target hybrid energy storage system, the output data and state of charge data of the target hybrid energy storage system during historical operation are retrieved. An anomaly detection algorithm is used to detect anomalies in the collected historical operation data of the target hybrid energy storage system. For example, by calculating the mean and standard deviation of the data, data points that exceed the threshold are identified as outliers. The threshold range is set according to expert experience and industry standards, such as the mean plus or minus two or three times the standard deviation. Through data anomaly detection, multiple historical energy storage anomaly data are obtained, forming a historical energy storage anomaly data set, including abnormal fluctuations in the output of the electrochemical energy storage unit or the gravity energy storage unit, and abnormal charging and discharging states of the energy storage unit, such as abnormal fluctuations in the state of charge or sudden capacity drops. The output data refers to the power output data generated by the energy storage unit during operation. Statistical analysis of the abnormal interval time in the historical energy storage anomaly data set determines the maximum abnormal interval time as the first collection bandwidth, and the minimum interval value as the second collection bandwidth.

[0030] According to the determined first collection bandwidth and second collection bandwidth, a first collection channel and a second collection channel are constructed. The first collection channel is used to collect data with a longer time scale and has a long collection frequency. The second collection channel is used to collect data with a shorter time scale and has a short collection frequency. Specifically, both the first collection channel and the second collection channel use a neural network model. Each collection channel includes a convolution layer, a pooling layer, and a normalization layer. The convolution layer is used for local time series feature extraction, the pooling layer is used to reduce feature dimension, and the normalization layer is used to eliminate data distribution differences and extract features from data with different sampling frequencies. The first collection channel and the second collection channel are connected in parallel, and their output layers are connected to their respective fully connected layers for data redundancy processing. The two collection channels are connected in parallel to form a dual-channel collection system that can process data with different time scales simultaneously, improving the efficiency and accuracy of data collection. Using the dual-channel collection system, the output data of the target hybrid energy storage system over a historical period is collected at different collection frequencies, and data redundancy processing is performed in the fully connected layer to obtain electrochemical energy output data sequences and gravity energy output data sequences. Through dual-channel collection, the output data sequences of electrochemical energy and gravity energy are obtained, improving the comprehensiveness and accuracy of the collected data.

[0031] Further, determining the first acquisition bandwidth and the second acquisition bandwidth based on the historical abnormal energy storage data set comprises: traversing to extract the abnormal interval time in the historical abnormal energy storage data set to obtain an abnormal interval time set; taking the maximum value in the abnormal interval time set as the first acquisition bandwidth; and taking the minimum value in the abnormal interval time set as the second acquisition bandwidth.

[0032] Specifically, the historical abnormal energy storage data set is traversed to extract the time interval between continuous abnormal data to form an abnormal interval time set, which contains multiple time intervals of multiple abnormal data of the target hybrid energy storage system in the historical operation process. The multiple abnormal time intervals in the abnormal interval time set are sorted in descending order, and according to the sorting, the maximum value in the abnormal interval time set is taken as the first acquisition bandwidth, and the slow trend characteristics of the output of the target hybrid energy storage system are acquired through a lower sampling frequency, and the minimum value in the abnormal interval time set is taken as the second acquisition bandwidth, and the rapid fluctuation characteristics of the output of the target hybrid energy storage system are acquired through a higher sampling frequency. By determining the first acquisition bandwidth and the second acquisition bandwidth according to the abnormal data time interval, different time scale sampling strategies are set for the historical abnormal energy storage data, and the comprehensiveness and accuracy of the data acquisition of the target hybrid energy storage system are improved.

[0033] Traversing the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to perform output trend prediction to obtain electrochemical energy storage output trend characteristics and gravity energy storage output trend characteristics.

[0034] Further, traversing the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to perform output trend prediction to obtain electrochemical energy storage output trend characteristics and gravity energy storage output trend characteristics comprises: using a multi-scale energy storage predictor to respectively perform output trend prediction on the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to obtain a multi-scale electrochemical energy storage output trend characteristic set and a multi-scale gravity energy storage output trend characteristic set; respectively performing leading interaction enhancement on the multi-scale electrochemical energy storage output trend characteristic set and the multi-scale gravity energy storage output trend characteristic set to determine a multi-scale enhanced electrochemical energy storage output trend characteristic set and a multi-scale enhanced gravity energy storage output trend characteristic set; and respectively calculating the mean value of the multi-scale enhanced electrochemical energy storage output trend characteristic set and the multi-scale enhanced gravity energy storage output trend characteristic set to obtain the electrochemical energy storage output trend characteristics and the gravity energy storage output trend characteristics.

[0035] Specifically, the electrochemical energy storage output data sequence and the gravity energy storage output data sequence are traversed, and the multi-scale energy storage predictor is used to perform prediction analysis on the electrochemical energy storage output data sequence and the gravity energy storage output data sequence respectively, to obtain a multi-scale electrochemical energy storage output trend feature set and a multi-scale gravity energy storage output trend feature set, each set containing features at multiple time scales, used to reflect the output trends of multiple energy storage units of the target hybrid energy storage system in different operating states.

[0036] The multi-scale energy storage predictor is a prediction model containing multiple different time scale analysis branches, each branch can independently process specific time scale data of the input energy storage output data sequence to obtain long-term trend, short-term fluctuation and periodic change features of the data. The multi-scale energy storage predictor includes an input layer, multiple time scale branches and an output layer, wherein the input layer inputs the historical energy storage output data sequence, including power data and state of charge time series. The time scale branch includes a long time scale branch and a short time scale branch, each branch uses a convolutional neural network or a time convolution network structure, different branches use different convolution kernel sizes or time steps to achieve multi-scale feature extraction. Finally, the output layer outputs a multi-scale output trend feature set, including predicted trend feature data at different time scales. In the training process of the multi-scale energy storage predictor, the historical output data of the target hybrid energy storage system is used as the training sample, the previous several time steps of the historical output sequence are used as the input, and the subsequent time trend is used as the label. The mean square error loss function is used for supervised training, and the gradient descent or Adam optimizer is used for iterative update of the multi-scale energy storage predictor parameters, so that the multi-scale energy storage predictor can accurately analyze the change features of the energy storage output of the target hybrid energy storage system at different time scales. After training, the trained multi-scale energy storage predictor is used to perform output trend prediction on the electrochemical energy storage output data sequence and the gravity energy storage output data sequence respectively.

[0037] The leading multi-scale electrochemical energy storage output trend feature set is interactively enhanced, and a leading multi-scale electrochemical energy storage output trend feature is extracted from the multi-scale electrochemical energy storage output trend feature set. The leading multi-scale electrochemical energy storage output trend feature is the feature with the highest overall similarity to other features in the multi-scale electrochemical energy storage output trend feature set. Based on the leading multi-scale electrochemical energy storage output trend feature, the enhancement of the multi-scale electrochemical energy storage output trend features in the multi-scale electrochemical energy storage output trend feature set is guided, and a multi-scale enhanced electrochemical energy storage output trend feature set is formed. At the same time, the leading multi-scale gravity energy storage output trend feature is extracted from the multi-scale gravity energy storage output trend feature set, and based on the leading multi-scale gravity energy storage output trend feature, the enhancement of the multi-scale gravity energy storage output trend features in the multi-scale gravity energy storage output trend feature set is guided, and a multi-scale enhanced gravity energy storage output trend feature set is formed.

[0038] Furthermore, the average value of the plurality of enhanced electrochemical energy storage output trend features in the multi-scale enhanced electrochemical energy storage output trend feature set is calculated to obtain an electrochemical energy storage output trend feature, which can reflect the output change trend of the electrochemical energy storage unit and provide reference data for the minute-level scheduling layer. By predicting the output change of the electrochemical energy storage, the charging and discharging strategy can be planned in advance to ensure that the load change can be quickly responded to when needed, and the frequency and voltage stability of the power grid can be maintained. At the same time, the average value of the plurality of enhanced gravity energy storage output trend features in the multi-scale enhanced gravity energy storage output trend feature set is calculated to obtain a gravity energy storage output trend feature, which is used to reflect the output change trend of the gravity energy storage unit and provide reference data for the hour-level scheduling layer.

[0039] By multi-scale trend prediction and leading interactive enhancement of the target hybrid energy storage system, the energy storage output characteristics of different time scales can be comprehensively obtained, high-reliable trend features can be provided for joint scheduling optimization of the target hybrid energy storage system, and the scheduling prediction accuracy, response speed, and new energy grid connection adaptability can be improved, so as to realize the collaborative optimization of fast response and large-scale energy support.

[0040] Further, the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are respectively subjected to leading interactive enhancement to determine a multi-scale enhanced electrochemical energy storage output trend feature set and a multi-scale enhanced gravity energy storage output trend feature set, including: extracting leading multi-scale electrochemical energy storage output trend features and leading multi-scale gravity energy storage output trend features from the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set; wherein the leading multi-scale electrochemical energy storage output trend features and the leading multi-scale gravity energy storage output trend features are respectively the features with the highest overall similarity to other multi-scale electrochemical energy storage output trend features and multi-scale gravity energy storage output trend features in the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set; based on the leading multi-scale electrochemical energy storage output trend features and the leading multi-scale gravity energy storage output trend features, the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are subjected to leading interactive enhancement to determine the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set.

[0041] Specifically, the multi-scale electrochemical energy storage output trend feature set is traversed, and a similarity calculation method such as cosine similarity, Euclidean distance, etc. is used to calculate the similarity between each feature in the multi-scale electrochemical energy storage output trend feature set and other features, and the sum of the similarity of each feature vector to other feature vectors in the set is calculated, and the multi-scale electrochemical energy storage output trend feature with the maximum similarity sum is extracted as the leading multi-scale electrochemical energy storage output trend feature. The maximum similarity sum reflects that the feature has the highest overall similarity to other multi-scale electrochemical energy storage output trend features in the multi-scale electrochemical energy storage output trend feature set, and can represent the main output feature trend of the multi-scale electrochemical energy storage output trend feature set. At the same time, the leading multi-scale gravity energy storage output trend feature is extracted from the multi-scale gravity energy storage output trend feature set through similarity calculation. The leading multi-scale gravity energy storage output trend feature is also the feature with the highest overall similarity to other multi-scale gravity energy storage output trend features in the multi-scale gravity energy storage output trend feature set.

[0042] Based on the screened leading multi-scale electrochemical energy storage output trend features and the leading multi-scale gravity energy storage output trend features, the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are respectively subjected to leading interactive enhancement. The interactive enhancement refers to using the leading features as parameters to weight and adjust other features in the set, so that the multiple features in the set maintain their own unique characteristics while being closer to the overall trend. Through leading interactive enhancement of the output, the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set are obtained.

[0043] By obtaining the leading feature and respectively performing the interactive enhancement mechanism on the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set based on the leading feature, the overall consistency of the feature set is improved, and the accuracy and stability of the energy storage output trend prediction are improved, more representative and stable input features are provided for the hierarchical joint of the target hybrid energy storage system, and the collaborative optimization of the hybrid energy storage system in rapid response and long-term energy management is realized.

[0044] Further, based on the leading multi-scale electrochemical energy storage output trend feature and the leading multi-scale gravity energy storage output trend feature, the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are interactively enhanced, and the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set are determined, including: calculating the fine-grained feature similarity of each multi-scale electrochemical energy storage output trend feature in the multi-scale electrochemical energy storage output trend feature set based on the leading multi-scale electrochemical energy storage output trend feature, obtaining a fine-grained similarity group set; determining an electrochemical leading interactive enhancement matrix set based on the fine-grained similarity group set; using the electrochemical leading interactive enhancement matrix set to perform leading enhancement on the multi-scale electrochemical energy storage output trend feature set, and obtaining the multi-scale enhanced electrochemical energy storage output trend feature set; based on the leading multi-scale gravity energy storage output trend feature, performing leading interactive enhancement on the multi-scale gravity energy storage output trend feature set, and determining the multi-scale enhanced gravity energy storage output trend feature set.

[0045] Specifically, for the multi-scale electrochemical energy storage output feature set, the fine-grained feature similarity between the leading multi-scale electrochemical energy storage output trend feature and each feature in the multi-scale electrochemical energy storage output feature set is calculated by using similarity calculation methods such as cosine similarity and Euclidean distance, the correlation between the leading multi-scale electrochemical energy storage output trend feature and other features in the multi-scale electrochemical energy storage output feature set is quantified, and the similarity between the two features is reflected. All similarity calculation results are integrated into a set to form a fine-grained similarity group set.

[0046] According to the calculated fine-grained similarity group set, the multiple similarity calculation results in the fine-grained similarity group set are normalized, the multiple similarity sums of the fine-grained similarity group set are 1, multiple feature weight vectors are formed, and each feature weight vector represents the similarity of the leading multi-scale electrochemical energy storage output trend feature to a certain feature. According to the multiple feature weight vectors, an electrochemical leading interactive enhancement matrix set is constructed, each row in the electrochemical leading interactive enhancement matrix set represents the weight distribution of the feature in the set, and each column corresponds to a time scale. The elements in the column represent the degree to which all candidate features are enhanced by the leading feature at that scale. The electrochemical leading interactive enhancement matrix set is used to lead and enhance the multi-scale electrochemical energy storage output trend feature set, that is, to perform weighted fusion. The multi-scale electrochemical energy storage output trend feature set and the leading multi-scale electrochemical energy storage output trend feature are combined according to the weights of the electrochemical leading interactive enhancement matrix set, and a multi-scale enhanced electrochemical energy storage output trend feature set is obtained. The multi-scale enhanced electrochemical energy storage output trend feature set can retain the independence of the features in the multi-scale electrochemical energy storage output trend feature set, and the consistency and representativeness of the whole multi-scale electrochemical energy storage output trend feature set are strengthened.

[0047] At the same time, the same method is used for the multi-scale gravity energy storage output trend feature set. First, the fine-grained similarity between the leading multi-scale gravity energy storage output trend feature and each feature in the multi-scale gravity energy storage output trend feature set is calculated to form a fine-grained similarity group set. According to the fine-grained similarity group set, a corresponding gravity leading interactive enhancement matrix set is constructed. Finally, the multi-scale gravity energy storage output trend feature set is led and interactively enhanced by the gravity leading interactive enhancement matrix set to determine a multi-scale enhanced gravity energy storage output trend feature set.

[0048] Through fine-grained calculation and enhancement matrix construction, the characteristics of the target hybrid energy storage system at different time scales are enhanced in detail, so that the determined multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set can more accurately reflect the output trend characteristics of the energy storage unit, thereby improving the accuracy and operation adaptability of the target hybrid energy storage system scheduling strategy, and achieving precise adaptation to grid multi-scenario operation requirements and efficient operation of the energy storage system.

[0049] The target hybrid energy storage system is monitored for operating state to obtain real-time system operating state monitoring data, and multi-objective hierarchical joint scheduling optimization is performed in combination with the electrochemical energy storage output trend feature and the gravity energy storage output trend feature to obtain a joint scheduling optimization scheme.

[0050] Specifically, the operation state of the target hybrid energy storage system is monitored and collected by the monitoring device and the data acquisition unit arranged, and real-time system operation state monitoring data is obtained. The monitoring device and the data acquisition unit include but are not limited to a current sensor, a voltage sensor, a power meter, a state monitoring module, etc. The current sensor and the voltage sensor are used to collect the charging and discharging current and voltage information of the electrochemical energy storage unit and the gravity energy storage unit in real time, so as to calculate the charging and discharging power of the energy storage unit. The power meter can directly measure the output power and input power of the energy storage unit. The state monitoring module includes but is not limited to a state of charge monitoring unit, a charging and discharging state monitoring unit, a shutdown state monitoring unit, and a hot standby state monitoring unit. The state of charge monitoring unit is used to obtain the remaining capacity or state of charge of the electrochemical energy storage unit in real time, to determine the current available energy of the electrochemical energy storage unit. The charging and discharging state monitoring unit is used to determine whether the energy storage unit is in a charging, discharging, or other state in real time. The shutdown state monitoring unit is used to monitor whether the gravity energy storage unit or the electrochemical energy storage unit is in a shutdown or maintenance state, to ensure that no command is issued to the shutdown unit during the dispatching optimization process. The hot standby state monitoring unit is used to monitor whether the energy storage unit is in a preheating or standby state, to ensure that the energy storage unit is enabled in time. The real-time system operation state monitoring data includes charging and discharging power, state of charge, charging and discharging state, shutdown state, and hot standby state, etc.

[0051] After obtaining the real-time system operation state monitoring data, the real-time system operation state monitoring data is subjected to multi-objective hierarchical joint dispatching optimization with the electrochemical energy storage output trend characteristics and the gravity energy storage output trend characteristics. The optimization objectives include reducing system operation cost and improving gravity energy storage consumption rate. Hierarchical joint dispatching means that the dispatching process is divided into different time scales according to the real-time system operation state monitoring data and the electrochemical energy storage output trend characteristics and the gravity energy storage output trend characteristics. The hour-level dispatching layer mainly faces the gravity energy storage, and utilizes its large capacity and slow response characteristics for energy balance and long-term support. The minute-level dispatching layer mainly faces the electrochemical energy storage, and utilizes the fast response characteristics of the electrochemical energy storage to realize frequency regulation and short-term balance. Through hierarchical joint dispatching, the advantages of different energy storage units of the target hybrid energy storage system are utilized, and complementary operation is realized. Through multi-objective hierarchical joint dispatching optimization, a joint dispatching optimization scheme is obtained. The joint dispatching optimization scheme includes the charging and discharging plan, the regulation strategy, and the operation constraint of each energy storage unit at different time scales.

[0052] Through real-time monitoring and trend characteristic fusion, the accuracy and adaptability of the hybrid energy storage system dispatching scheme are improved, the efficient cooperation of different types of energy storage units in the hybrid energy storage system at multiple time scales is ensured, the reliability and stability of the hybrid energy storage system operation are improved, the stability of the grid frequency and voltage is ensured, and the fluctuation is reduced.

[0053] Further, the scheduling optimization multi-objective is to reduce system operation cost and improve gravity energy storage consumption rate.

[0054] Further, the scheduling optimization multi-objective is to reduce system operation cost and improve gravity energy storage consumption rate.

[0055] Specifically, the scheduling optimization multi-objective is to not only meet the basic requirement of power supply and demand balance in the scheduling process, but also to reduce system operation cost and improve gravity energy storage consumption rate. The system operation cost refers to minimizing the operation cost of the hybrid energy storage system, reducing the high-frequency charge and discharge loss of the electrochemical energy storage unit. The improvement of the gravity energy storage consumption rate refers to maximizing the utilization rate of the gravity energy storage, avoiding its capacity idling, and improving the resource utilization efficiency. The scheduling optimization multi-objective is used as a constraint, and the real-time system operation state monitoring data, the electrochemical energy storage output trend characteristics and the gravity energy storage output trend characteristics are used as input data, which are input into the joint scheduling optimization layer. The joint scheduling optimization layer includes a hourly gravity energy storage scheduling layer and a minute-level electrochemical energy storage scheduling layer. The hourly gravity energy storage scheduling layer is mainly used to process scheduling tasks of a longer time scale. Considering that the gravity energy storage unit usually has a large energy storage capacity and a long charge and discharge period, the scheduling decision of the hourly gravity energy storage scheduling layer is how to reasonably arrange the charge and discharge of the gravity energy storage in a longer time window to improve its consumption rate, that is, to fully utilize the gravity energy storage to meet the power grid load demand while avoiding excessive charge and discharge of the gravity energy storage. The minute-level electrochemical energy storage scheduling layer is used to process scheduling tasks of a shorter time scale. The electrochemical energy storage unit has fast response characteristics and can provide or absorb a large amount of power in a short time. The scheduling decision of the minute-level electrochemical energy storage scheduling layer focuses on how to utilize the electrochemical energy storage to quickly respond to the instantaneous changes of the load, maintain the frequency and voltage stability of the power grid, and at the same time consider the charge and discharge state and health condition of the electrochemical energy storage unit to reduce the operation cost of the electrochemical energy storage unit.

[0056] The joint scheduling optimization layer, based on real-time system operation status monitoring data, electrochemical energy storage output trend characteristics, and gravity energy storage output trend characteristics, considers two multi-objective constraints: reducing system operating costs and improving gravity energy storage absorption rate. It uses an optimization algorithm to find the optimal scheduling scheme and outputs a joint scheduling optimization scheme. The optimization algorithm can employ multi-objective particle swarm optimization or a multi-objective genetic algorithm. For example, in the multi-objective hierarchical joint scheduling optimization process, hourly gravity energy storage output and minute-level electrochemical energy storage output are used as decision traversals, minimizing system operating costs and maximizing gravity energy storage absorption rate are used as objective functions, and basic conditions such as energy storage unit capacity limitations, charge / discharge rate limitations, battery state of charge, and grid power balance are used as constraints input into the multi-objective genetic optimization algorithm. The multi-objective genetic optimization algorithm outputs multiple candidate joint scheduling optimization schemes based on real-time system operation status data, the predicted trend characteristics of the two energy storage outputs, and the multi-objective constraints. Each scheme corresponds to the charge / discharge power allocation of different energy storage units in the target hybrid energy storage system at different time scales, and calculates the objective function value and constraint satisfaction. After obtaining multiple candidate joint scheduling optimization schemes, the multi-objective genetic optimization algorithm further updates these schemes iteratively through selection, crossover, and mutation operations to generate a new generation of schemes. In each iteration, the top K joint scheduling optimization schemes that satisfy the constraints are retained, forming the Pareto front solution set. During hierarchical scheduling, the hourly gravity energy storage scheduling results are used as boundary conditions input to the minute-level electrochemical energy storage scheduling layer to ensure consistent outputs. Simultaneously, the power allocation between the two levels is dynamically adjusted in each iteration to satisfy multi-objective constraints. Finally, when the multi-objective genetic optimization algorithm converges or reaches the maximum iteration limit, the optimal scheme is selected from the Pareto front schemes to form a joint scheduling optimization scheme, including the charging and discharging power allocation and adjustment strategies for each energy storage unit at the hourly and minute levels.

[0057] The joint scheduling optimization scheme can be directly used in the control system or scheduling platform of the target hybrid energy storage system to realize the dynamic coordinated operation of the target hybrid energy storage system. While ensuring that multiple objective constraints are met, it can realize the complementary coordination of different energy storage units, improve the real-time performance and adaptability of the scheduling scheme, and thus enhance the stability and reliability of the power grid operation.

[0058] Example 2, based on the same inventive concept as the hierarchical joint scheduling method for a hybrid energy storage system in the foregoing examples, such as... Figure 2 As shown, this application provides a hierarchical joint scheduling system for hybrid energy storage systems, wherein the hierarchical joint scheduling system for hybrid energy storage systems includes:

[0059] The data acquisition module 11 is configured to acquire double-channel data of the output data of the target hybrid energy storage system in a historical time, and obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence; the trend prediction module 12 is configured to traverse the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to predict the output trend, and obtain an electrochemical energy storage output trend feature and a gravity energy storage output trend feature; and the scheduling optimization module 13 is configured to monitor the running state of the target hybrid energy storage system, obtain real-time system running state monitoring data, and combine the electrochemical energy storage output trend feature and the gravity energy storage output trend feature to perform multi-objective hierarchical joint scheduling optimization, and obtain a joint scheduling optimization scheme.

[0060] Further, the data acquisition module 11 is further configured to: acquire a historical energy storage abnormal data set of the target hybrid energy storage system, determine a first acquisition bandwidth and a second acquisition bandwidth based on the historical energy storage abnormal data set; construct a first acquisition channel and a second acquisition channel based on the first acquisition bandwidth and the second acquisition bandwidth respectively; obtain an acquisition double channel by connecting the first acquisition channel and the second acquisition channel in parallel, and connecting output layers to full connection layers respectively; and acquire output data of the target hybrid energy storage system in a historical time by using the acquisition double channel, and obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence.

[0061] Further, the data acquisition module 11 is further configured to: traverse and extract abnormal interval times in the historical energy storage abnormal data set to obtain an abnormal interval time set; take a maximum value in the abnormal interval time set as the first acquisition bandwidth; and take a minimum value in the abnormal interval time set as the second acquisition bandwidth.

[0062] Further, the trend prediction module 12 is further configured to: use a multi-scale energy storage predictor to respectively predict output trends of the electrochemical energy storage output data sequence and the gravity energy storage output data sequence, and obtain a multi-scale electrochemical energy storage output trend feature set and a multi-scale gravity energy storage output trend feature set; perform leading interaction enhancement on the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set respectively, and determine a multi-scale enhanced electrochemical energy storage output trend feature set and a multi-scale enhanced gravity energy storage output trend feature set; and calculate the mean values of the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set respectively, and obtain the electrochemical energy storage output trend feature and the gravity energy storage output trend feature.

[0063] Further, the trend prediction module 12 is further configured to: filter and extract a leading multi-scale electrochemical energy storage output trend feature and a leading multi-scale gravity energy storage output trend feature from the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set; wherein the leading multi-scale electrochemical energy storage output trend feature and the leading multi-scale gravity energy storage output trend feature are respectively the features with the highest overall similarity to other multi-scale electrochemical energy storage output trend features and multi-scale gravity energy storage output trend features in the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set; and perform leading interaction enhancement on the multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set based on the leading multi-scale electrochemical energy storage output trend feature and the leading multi-scale gravity energy storage output trend feature, to determine the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set.

[0064] Further, the trend prediction module 12 is further configured to: calculate a fine-grained feature similarity of the leading multi-scale electrochemical energy storage output trend feature to each multi-scale electrochemical energy storage output trend feature in the multi-scale electrochemical energy storage output trend feature set, to obtain a fine-grained similarity group set; determine an electrochemical leading interaction enhancement matrix set based on the fine-grained similarity group set; perform leading enhancement on the multi-scale electrochemical energy storage output trend feature set by using the electrochemical leading interaction enhancement matrix set, to obtain the multi-scale enhanced electrochemical energy storage output trend feature set; and perform leading interaction enhancement on the multi-scale gravity energy storage output trend feature set based on the leading multi-scale gravity energy storage output trend feature, to determine the multi-scale enhanced gravity energy storage output trend feature set.

[0065] Further, the scheduling optimization module 13 is further configured to: perform hierarchical optimization based on the real-time system operation state monitoring data, the electrochemical energy storage output trend feature, and the gravity energy storage output trend feature by using a joint scheduling optimization layer, to obtain a joint scheduling optimization scheme, with a scheduling optimization multi-objective as a constraint; wherein the joint scheduling optimization layer includes a gravity energy storage scheduling layer at an hour level and an electrochemical energy storage scheduling layer at a minute level.

[0066] Further, the scheduling optimization module 13 is further configured to: the scheduling optimization multi-objective is to reduce system operation cost and improve gravity energy storage consumption rate.

[0067] In the third embodiment, based on the same inventive concept as the hierarchical joint scheduling method of the hybrid energy storage system in the foregoing embodiments, the present application also provides a computer readable storage medium, and the computer readable storage medium stores a computer program. When the computer program is executed, the steps of the hierarchical joint scheduling method of the hybrid energy storage system in any one of the foregoing embodiments are implemented.

[0068] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Numerous modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0069] Obviously, many modifications and changes can be made to the present application by those skilled in the art which do not depart from the spirit and scope of the application.

Claims

1. A hierarchical co-scheduling method for a hybrid energy storage system, characterized in that, The method includes: Dual-channel acquisition was performed on the output data of the target hybrid energy storage system over a historical period to obtain electrochemical energy storage output data sequences and gravity energy storage output data sequences. The output trend is predicted by traversing the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to obtain the output trend characteristics of electrochemical energy storage and gravity energy storage. The target hybrid energy storage system is monitored for operation status to obtain real-time system operation status monitoring data. Multi-objective hierarchical joint scheduling optimization is performed by combining the output trend characteristics of electrochemical energy storage and gravity energy storage to obtain a joint scheduling optimization scheme. The output trend is predicted by traversing the electrochemical energy storage output data sequence and the gravity energy storage output data sequence to obtain the output trend characteristics of electrochemical energy storage and gravity energy storage, including: The output trend of the electrochemical energy storage output data sequence and the gravity energy storage output data sequence are predicted by a multi-scale energy storage predictor, respectively, to obtain a multi-scale electrochemical energy storage output trend feature set and a multi-scale gravity energy storage output trend feature set. The multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are respectively guided and enhanced to determine the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set. The mean values ​​of the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set are calculated respectively to obtain the electrochemical energy storage output trend features and the gravity energy storage output trend features. The multi-scale electrochemical energy storage output trend feature set and the multi-scale gravity energy storage output trend feature set are respectively subjected to guided interactive enhancement to determine the multi-scale enhanced electrochemical energy storage output trend feature set and the multi-scale enhanced gravity energy storage output trend feature set, including: The characteristics leading the multi-scale electrochemical energy storage output trend and the characteristics leading the multi-scale gravity energy storage output trend are selected and extracted from the set of multi-scale electrochemical energy storage output trend and the set of multi-scale gravity energy storage output trend. Among them, the features that lead the multi-scale electrochemical energy storage output trend and the features that lead the multi-scale gravity energy storage output trend are the features with the highest overall similarity to other multi-scale electrochemical energy storage output trend features and multi-scale gravity energy storage output trend features, respectively. Based on the aforementioned leading multi-scale electrochemical energy storage output trend characteristics and leading multi-scale gravity energy storage output trend characteristics, a leading interactive enhancement is performed on the multi-scale electrochemical energy storage output trend characteristic set and the multi-scale gravity energy storage output trend characteristic set to determine the multi-scale enhanced electrochemical energy storage output trend characteristic set and the multi-scale enhanced gravity energy storage output trend characteristic set, including: Calculate the fine-grained feature similarity between the leading multi-scale electrochemical energy storage output trend feature and each multi-scale electrochemical energy storage output trend feature in the multi-scale electrochemical energy storage output trend feature set, and obtain a fine-grained similarity set. determining an electrochemical leading interaction enhancement matrix set based on the fine-grained similarity group set; performing leading enhancement on the multi-scale electrochemical energy storage output trend feature set by using the electrochemical leading interaction enhancement matrix set, to obtain a multi-scale enhanced electrochemical energy storage output trend feature set; performing leading interaction enhancement on the multi-scale gravity energy storage output trend feature set based on the leading multi-scale gravity energy storage output trend feature, to determine a multi-scale enhanced gravity energy storage output trend feature set.

2. The hierarchical co-scheduling method of a hybrid energy storage system according to claim 1, wherein, performing double-channel collection on output data of a target hybrid energy storage system in a historical time, to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence, including: obtaining a historical energy storage abnormal data set of the target hybrid energy storage system, and determining a first collection bandwidth and a second collection bandwidth based on the historical energy storage abnormal data set; constructing a first collection channel and a second collection channel based on the first collection bandwidth and the second collection bandwidth respectively; connecting the first collection channel and the second collection channel in parallel, and connecting output layers to fully connected layers respectively, to obtain a collection double channel; collecting output data of the target hybrid energy storage system in a historical time by using the collection double channel, to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence.

3. The hierarchical co-scheduling method of claim 2, wherein, determining a first collection bandwidth and a second collection bandwidth based on the historical energy storage abnormal data set, including: traversing and extracting abnormal interval times in the historical energy storage abnormal data set, to obtain an abnormal interval time set; taking a maximum value in the abnormal interval time set as the first collection bandwidth; taking a minimum value in the abnormal interval time set as the second collection bandwidth.

4. The hierarchical co-scheduling method of a hybrid energy storage system according to claim 1, wherein, performing real-time system running state monitoring on the target hybrid energy storage system, to obtain real-time system running state monitoring data, and performing multi-target hierarchical joint scheduling optimization on the electrochemical energy storage output trend feature and the gravity energy storage output trend feature, to obtain a joint scheduling optimization scheme, including: performing hierarchical optimization on the real-time system running state monitoring data, the electrochemical energy storage output trend feature and the gravity energy storage output trend feature based on the joint scheduling optimization layer, to obtain a joint scheduling optimization scheme, with the scheduling optimization multi-target as a constraint; wherein the joint scheduling optimization layer includes a gravity energy storage scheduling layer at an hourly level and an electrochemical energy storage scheduling layer at a minute level.

5. The hierarchical co-scheduling method of a hybrid energy storage system according to claim 4, wherein, The scheduling optimization multi-target is to reduce system running cost and improve gravity energy storage consumption rate.

6. A hybrid energy storage system layered co-scheduling system, characterized in that, a data collection module for performing double-channel collection on output data of a target hybrid energy storage system in a historical time, to obtain an electrochemical energy storage output data sequence and a gravity energy storage output data sequence; a trend prediction module for performing output trend prediction on the electrochemical energy storage output data sequence and the gravity energy storage output data sequence by traversal, to obtain an electrochemical energy storage output trend feature and a gravity energy storage output trend feature; ​ The scheduling optimization module is configured to perform operation state monitoring on the target hybrid energy storage system, obtain real-time system operation state monitoring data, perform multi-objective hierarchical joint scheduling optimization in combination with the electrochemical energy storage output trend characteristics and the gravity energy storage output trend characteristics, and obtain a joint scheduling optimization scheme.

7. A computer-readable storage medium, characterized in that, The computer program is stored on the computer readable storage medium and is configured to implement the steps of the hybrid energy storage system hierarchical joint scheduling method according to any one of claims 1 to 5 when executed.

Images