A peak load shifting optimization scheduling method for distributed energy storage power station in photovoltaic grid-connected time

Abstract

The application relates to the technical field of electric energy storage, and particularly discloses a distributed energy storage power station peak shaving and valley filling optimization scheduling method during photovoltaic grid connection, which comprises the following steps: performing time sequence deduction on the net load power of a photovoltaic energy storage power station to obtain the net load change trend of the photovoltaic energy storage power station; performing state membership discrimination on the current state of charge value and the health state value of the photovoltaic energy storage power station to obtain a dynamic priority queue of the photovoltaic energy storage power station; determining the scheduling mode of the photovoltaic energy storage power station and performing corresponding scheduling processes of the scheduling mode; adaptively correcting the dynamic priority queue and rolling optimizing the subsequent scheduling sequence after correction of the dynamic priority queue; when the state of charge of any distributed energy storage node reaches a preset charging and discharging threshold value, temporarily removing the distributed energy storage node from the dynamic priority queue and triggering the balancing protection action of the distributed energy storage node; and the application can improve the efficiency of the distributed energy storage power station peak shaving and valley filling optimization scheduling.

Description

Technical Field

[0001] This invention relates to the field of energy storage technology, and in particular to a peak shaving and valley filling optimization scheduling method for distributed energy storage power stations connected to the photovoltaic grid. Background Technology

[0002] In existing photovoltaic grid-connected scenarios, the peak shaving and valley filling scheduling of distributed energy storage power stations only adopts fixed scheduling rules for operation and management. It cannot rely on the historical status records of the power station to complete the time-series prediction of net load power, cannot accurately form the trend of net load changes, and lacks reliable data support for scheduling decisions.

[0003] The existing scheduling method does not combine the charge state and health state of distributed energy storage nodes to determine the state of membership, and cannot generate a dynamic priority queue that adapts to the real-time operating status of the power station. During the scheduling process, it cannot modify the queue and optimize the scheduling order based on the real-time status of the nodes, and it does not set up a balance protection mechanism triggered by the node charging and discharging threshold. As a result, the overall scheduling efficiency is low and the power station operation stability is insufficient. Summary of the Invention

[0004] This invention provides a peak shaving and valley filling optimization scheduling method for distributed energy storage power stations when photovoltaic grid is connected, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection, comprising: Itm1. Based on the historical status records of photovoltaic energy storage power stations, the net load power of photovoltaic energy storage power stations is extrapolated over time to obtain the net load change trend of photovoltaic energy storage power stations; Itm2. Determine the state membership of the photovoltaic energy storage power station based on its current state of charge and health status values ​​to obtain the dynamic priority queue of the photovoltaic energy storage power station; Itm3. Based on the net load change trend, determine the dispatch mode of the photovoltaic energy storage power station and execute the corresponding dispatch process of the dispatch mode; Itm4. During the scheduling process, the dynamic priority queue is adaptively corrected based on the real-time charge status and health status of the distributed energy storage nodes, and the corrected dynamic priority queue is then used to optimize the subsequent scheduling order. Itm5. During the current scheduling cycle of the photovoltaic energy storage power station, when the state of charge of any distributed energy storage node reaches the preset charging and discharging threshold, the distributed energy storage node will be temporarily removed from the dynamic priority queue, and the balancing protection action of the distributed energy storage node will be triggered.

[0006] In a preferred embodiment, the step of performing time-series extrapolation of the net load power of the photovoltaic energy storage power station based on its historical state records to obtain the net load change trend of the photovoltaic energy storage power station includes: Read the historical net load sequence from the historical operation database of the photovoltaic energy storage power station; By performing a collaborative analysis of the net load feature vector in the historical net load sequence and the current net load sequence of the photovoltaic energy storage power station, the net load change trend of the photovoltaic energy storage power station can be obtained.

[0007] In a preferred embodiment, the step of determining the state membership of the current state of charge and health status values ​​of the photovoltaic energy storage power station to obtain a dynamic priority queue for the photovoltaic energy storage power station includes: Based on the charging and discharging threshold range of the photovoltaic energy storage power station, determine the charge state affiliation label and health state affiliation label of the distributed energy storage nodes in the photovoltaic energy storage power station. Combine the charge state affiliation tag and the health state affiliation tag into a two-dimensional affiliation tag pair; Fuzzy reasoning is performed on the two-dimensional membership label pairs to obtain the dynamic priority queue of the photovoltaic energy storage power station.

[0008] In a preferred embodiment, the step of performing fuzzy reasoning on the two-dimensional membership label pairs to obtain the dynamic priority queue of the photovoltaic energy storage power station includes: The charge state membership label and health state membership label of the distributed energy storage node are input into the fuzzy inference rule base of the photovoltaic energy storage power station for matching calculation to obtain the priority value of the two-dimensional membership label pair. A dynamic priority queue for photovoltaic energy storage power stations is constructed based on priority values.

[0009] In a preferred embodiment, the priority value is calculated using the following formula: ; In the formula, This represents the priority value of the distributed energy storage node. This represents the total number of rules in the fuzzy reasoning rule base. For the first The matching degree of the antecedent of the charge state under the rule, For the first The health status antecedent matching degree of the rule. The preset semantic distance attenuation coefficient, For the first The comprehensive semantic distance of the rules, For the first The priority weight of the consequent of a rule. It is a natural constant.

[0010] In a preferred embodiment, determining the dispatch mode of the photovoltaic energy storage power station based on the net load change trend and executing the corresponding dispatch process for the dispatch mode includes: By comparing and analyzing the current net load value in the net load change trend with the reference value of the upper limit of allowable feed-in at the grid connection point of the photovoltaic energy storage power station, the grid connection point value range status of the net load change trend is obtained. Based on the changing pattern of net load, determine the trend characteristics of net load change. Based on the value range status and trend characteristics of the grid connection point, the dispatch mode of the photovoltaic energy storage power station is determined. The dispatch modes include peak shaving mode, valley filling mode and maintenance mode. In peak shaving mode, distributed energy storage nodes are selected sequentially from high priority to low priority according to the dynamic priority queue to perform discharge operations until the current net load value drops below the upper limit reference value allowed for grid connection. In valley filling mode, distributed energy storage nodes are selected sequentially from low priority to high priority according to the dynamic priority queue to perform charging operations until the current net load value rises above the lower limit reference value allowed by the grid connection point.

[0011] In a preferred embodiment, the step of adaptively correcting the dynamic priority queue based on the real-time state of charge and health status of the distributed energy storage nodes during the scheduling process, and then using the corrected dynamic priority queue to continuously optimize the subsequent scheduling order, includes: After the scheduling step of the photovoltaic energy storage power station is completed, the real-time state of charge is updated by superimposing the real-time state of charge change of the distributed energy storage nodes. Using the health status of distributed energy storage nodes as fixed attribute values, the updated state of charge is re-executed with state membership determination and fuzzy reasoning to obtain the corrected priority label of the updated state of charge. Based on the priority label, the dynamic priority queue is adaptively corrected.

[0012] In a preferred embodiment, the rolling optimization of the subsequent scheduling order includes: The modified priority label will be used as the basis for scheduling distributed energy storage nodes in the next scheduling step. The distributed energy storage nodes are reordered according to the priority values ​​in the corrected priority labels, and the corrected dynamic priority queue is used as the input for the next scheduling step. The complete cycle of scheduling mode determination, charging and discharging operation execution, state of charge acquisition, and priority queue correction is repeated until the current scheduling cycle of the photovoltaic energy storage power station ends.

[0013] In a preferred embodiment, the step of, within the current scheduling cycle of the photovoltaic energy storage power station, when the state of charge of any distributed energy storage node reaches a preset charging / discharging threshold, includes: The real-time state of charge (SOC) values ​​of distributed energy storage nodes in the dynamic priority queue are traversed at fixed time intervals. The real-time state of charge value is compared with the preset charge and discharge threshold in real time to obtain the state of charge range of the real-time state of charge value.

[0014] In a preferred embodiment, temporarily removing the distributed energy storage node from the dynamic priority queue and triggering the distributed energy storage node's load balancing protection action includes: When the current state of charge value read is within the normal range of the state of charge, the original position in the dynamic priority queue remains unchanged. When the current state of charge value read deviates from the normal range of the state of charge, it is recorded as an abnormal state node. Store the node identifier and current state of charge value of the abnormal node in the temporary monitoring record table, and suspend the participation qualification of the abnormal node in subsequent scheduling instructions until the state of charge of the abnormal node returns to the normal state of charge range.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention uses historical state records of photovoltaic energy storage power stations to perform time-series extrapolation of net load power, accurately generating net load change trends and providing stable data support for scheduling decisions, making scheduling mode determination more aligned with the actual operational needs of the power station. By performing state affiliation judgment on the charge and health states of distributed energy storage nodes and constructing a dynamic priority queue, the selection of scheduling targets becomes more targeted, significantly improving the accuracy of scheduling execution.

[0016] 2. This invention adaptively adjusts the dynamic priority queue based on the real-time status of nodes during the scheduling process, and continuously optimizes the subsequent scheduling order to ensure that the scheduling process continuously adapts to changes in the operating status of the power station. By triggering node balancing protection actions through preset charging and discharging thresholds, the charging and discharging behavior of distributed energy storage nodes is standardized, comprehensively improving the overall efficiency of peak shaving and valley filling scheduling of distributed energy storage power stations during photovoltaic grid connection, while enhancing the stability and safety of power station operation. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a peak-shaving and valley-filling optimization scheduling method for a distributed energy storage power station during photovoltaic grid connection, provided in an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection. The execution entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a peak-shaving and valley-filling optimization scheduling method for a distributed energy storage power station during photovoltaic grid connection, according to an embodiment of the present invention. In this embodiment, the peak-shaving and valley-filling optimization scheduling method for a distributed energy storage power station during photovoltaic grid connection includes: Itm1. Based on the historical status records of photovoltaic energy storage power stations, the net load power of photovoltaic energy storage power stations is extrapolated over time to obtain the net load change trend of photovoltaic energy storage power stations; In this embodiment of the invention, the step of performing time-series extrapolation of the net load power of the photovoltaic energy storage power station based on its historical state records to obtain the net load change trend of the photovoltaic energy storage power station includes: Read the historical net load sequence from the historical operation database of the photovoltaic energy storage power station; By performing a collaborative analysis of the net load feature vector in the historical net load sequence and the current net load sequence of the photovoltaic energy storage power station, the net load change trend of the photovoltaic energy storage power station can be obtained.

[0021] From the dedicated historical operation database pre-built and continuously storing operation data of the photovoltaic energy storage power station, the historical data retrieval time range is set according to the past continuous operation time range pre-set by the power station before startup and scheduling. The net load data corresponding to all collection nodes within this time range is completely extracted according to the time storage order of the data in the database. All historical net load sequences stored in the database are read without any data omissions or truncation. The historical net load sequence accurately records the measured net load value corresponding to each fixed collection time node in the past operation cycle of the power station, and completely covers all net load data information in the corresponding time period.

[0022] The system extracts operational characteristic information such as time period attributes, numerical fluctuation status, and load stability for each set of measured net load values ​​in the historical net load sequence. All extracted characteristic information is integrated into a net load feature vector according to fixed rules. This net load feature vector is then compared with the current net load sequence generated in real time at the same fixed acquisition time node under the current operating state of the photovoltaic energy storage power station. Point-by-point numerical matching and feature correlation analysis are performed on each set of corresponding acquisition time nodes. The feature matching results and correlation patterns of all time nodes are summarized and integrated to form a complete and continuous net load change trend of the photovoltaic energy storage power station.

[0023] Itm2. Determine the state membership of the photovoltaic energy storage power station based on its current state of charge and health status values ​​to obtain the dynamic priority queue of the photovoltaic energy storage power station; In this embodiment of the invention, the step of determining the state membership of the current state of charge and health status of the photovoltaic energy storage power station to obtain the dynamic priority queue of the photovoltaic energy storage power station includes: Based on the charging and discharging threshold range of the photovoltaic energy storage power station, determine the charge state affiliation label and health state affiliation label of the distributed energy storage nodes in the photovoltaic energy storage power station. Combine the charge state affiliation tag and the health state affiliation tag into a two-dimensional affiliation tag pair; Fuzzy reasoning is performed on the two-dimensional membership label pairs to obtain the dynamic priority queue of the photovoltaic energy storage power station.

[0024] The process of performing fuzzy reasoning on two-dimensional membership label pairs to obtain the dynamic priority queue of the photovoltaic energy storage power station includes: The charge state membership label and health state membership label of the distributed energy storage node are input into the fuzzy inference rule base of the photovoltaic energy storage power station for matching calculation to obtain the priority value of the two-dimensional membership label pair. A dynamic priority queue for photovoltaic energy storage power stations is constructed based on priority values.

[0025] The formula for calculating the priority value is as follows: ; In the formula, This represents the priority value of the distributed energy storage node. This represents the total number of rules in the fuzzy reasoning rule base. For the first The matching degree of the antecedent of the charge state under the rule, For the first The health status antecedent matching degree of the rule. The preset semantic distance attenuation coefficient, For the first The comprehensive semantic distance of the rules, For the first The priority weight of the consequent of a rule. It is a natural constant.

[0026] Based on the pre-set charging and discharging threshold ranges of the photovoltaic energy storage power station according to the hardware parameters and safe operation specifications of the distributed energy storage nodes before commissioning, this range includes the upper limit of charging threshold, the lower limit of discharging threshold, and the normal operating range of the charged state. The current state of charge value collected in real time for each distributed energy storage node is compared with this charging and discharging threshold range to accurately determine the threshold sub-range in which the current state of charge value belongs, thus determining the unique state of charge affiliation label for that distributed energy storage node. Simultaneously, based on the pre-set health status judgment threshold ranges of the photovoltaic energy storage power station according to the battery life, degradation degree, and operating performance, this range includes the healthy operation range, the sub-health warning range, and the abnormal fault range. The current health status value detected in real time by the distributed energy storage node is compared with this health status judgment threshold range to accurately determine the threshold sub-range in which the current health status value belongs, thus determining the unique health status affiliation label for that distributed energy storage node.

[0027] The state of charge (SCC) and health status (HS) labels of each independent distributed energy storage node are paired and integrated one-to-one according to a pre-defined fixed combination order of SCC label first and HHS label second, forming a unique two-dimensional label pair that belongs only to that distributed energy storage node and is not repeated with other nodes.

[0028] All the two-dimensional membership tag pairs corresponding to the distributed energy storage nodes are sequentially matched with all the judgment rules in the fuzzy inference rule base that is pre-built and fixed in the scheduling system of the photovoltaic energy storage power station, according to the node number order. Each judgment rule in the fuzzy inference rule base has a preset correspondence between the two-dimensional membership tag combination and the priority level. Based on the judgment rule that successfully matches the node's two-dimensional membership tag pair, the priority level corresponding to each distributed energy storage node is directly determined. Then, all distributed energy storage nodes are arranged in order according to the fixed arrangement rule from high level to low level, and finally a dynamic priority queue of the photovoltaic energy storage power station that is adapted to the real-time operation status of the power station is formed.

[0029] The state of charge (SOC) and health status (HS) labels corresponding to a single distributed energy storage node are sequentially sent into a pre-defined fuzzy inference rule base within the photovoltaic energy storage power station dispatch system, according to the order in which the labels are generated. The core feature information of the SOC and HHS labels is then matched and filtered item by item according to all the pre-written judgment rules in the fuzzy inference rule base. Based on all the matching judgment rules, the corresponding results are summarized, integrated, and determined, ultimately yielding the priority value corresponding to the two-dimensional SOC label pair of the distributed energy storage node.

[0030] Based on the priority values ​​calculated by each distributed energy storage node, all distributed energy storage nodes are uniformly sorted using a fixed sorting method from nodes with higher values ​​to nodes with lower values. After sorting, all distributed energy storage node combinations are standardized and organized, duplicate node information is removed, node identification information is supplemented, and finally a dynamic priority queue of the photovoltaic energy storage power station is fully constructed.

[0031] The total number of rules used in the priority value calculation process is obtained by counting and statistically analyzing all the preset judgment rules stored in the fuzzy inference rule base of the photovoltaic energy storage power station, and the accurate total number of rules is used as the basis for calculation.

[0032] The state of charge antecedent matching degree is obtained by comparing the state of charge of the distributed energy storage node with the antecedent conditions of a single decision rule in the fuzzy inference rule base item by item to determine the degree of complete fit between the label and the rule antecedent conditions.

[0033] The health status antecedent matching degree is obtained by comparing the health status affiliation label of the distributed energy storage node with the antecedent conditions of a single judgment rule in the fuzzy inference rule base item by item to determine the degree of complete fit between the label and the rule antecedent conditions.

[0034] The semantic distance attenuation coefficient is a fixed standard value that is pre-set before the photovoltaic energy storage power station is officially started and dispatched, taking into account the operating characteristics of distributed energy storage nodes and dispatch response requirements. This value remains constant throughout the entire dispatch cycle.

[0035] The comprehensive semantic distance is obtained by measuring the differences between the overall features of the distributed energy storage nodes and the antecedent features of individual judgment rules within the fuzzy inference rule base through the two-dimensional membership labels of the nodes. This accurately quantifies the degree of semantic difference between the two and obtains the corresponding comprehensive semantic distance.

[0036] The priority weight of the consequent is a fixed weight value that is pre-configured in the rule writing stage of each judgment rule in the fuzzy inference rule base of the photovoltaic energy storage power station. This value remains fixed and is not adjusted after the rule base is enabled.

[0037] The natural constant adopts a fixed standard value that is universally accepted and established in the field of mathematics, and is directly substituted into the overall calculation process of the priority value to participate in the complete calculation.

[0038] The priority numerical calculation process involves performing sequential operations on the feature information obtained from the matching of the judgment rules, the calculated semantic distance value, and the pre-set weight information according to the preset integration logic, and finally obtaining a precise quantitative result that can directly represent the scheduling priority of distributed energy storage nodes.

[0039] The priority value is specifically used to provide a unique and exclusive sorting criterion for each distributed energy storage node. This value directly serves as core data to support the complete construction of the dynamic priority queue of the photovoltaic energy storage power station.

[0040] Priority values ​​can accurately distinguish the scheduling order of different distributed energy storage nodes, ensuring that the dynamic priority queue arrangement always adapts to the real-time changing operating status of the photovoltaic energy storage power station, thus guaranteeing the rationality and adaptability of the scheduling order.

[0041] Itm3. Based on the net load change trend, determine the dispatch mode of the photovoltaic energy storage power station and execute the corresponding dispatch process of the dispatch mode; In this embodiment of the invention, determining the scheduling mode of the photovoltaic energy storage power station based on the net load change trend and executing the corresponding scheduling process of the scheduling mode includes: By comparing and analyzing the current net load value in the net load change trend with the reference value of the upper limit of allowable feed-in at the grid connection point of the photovoltaic energy storage power station, the grid connection point value range status of the net load change trend is obtained. Based on the changing pattern of net load, determine the trend characteristics of net load change. Based on the value range status and trend characteristics of the grid connection point, the dispatch mode of the photovoltaic energy storage power station is determined. The dispatch modes include peak shaving mode, valley filling mode and maintenance mode. In peak shaving mode, distributed energy storage nodes are selected sequentially from high priority to low priority according to the dynamic priority queue to perform discharge operations until the current net load value drops below the upper limit reference value allowed for grid connection. In valley filling mode, distributed energy storage nodes are selected sequentially from low priority to high priority according to the dynamic priority queue to perform charging operations until the current net load value rises above the lower limit reference value allowed by the grid connection point.

[0042] The current net load value collected in real time in the net load change trend is compared with the reference value of the upper limit of allowable feed into the grid connection point, which is set in advance by the photovoltaic energy storage power station according to the grid connection safety operation specifications and grid access requirements. The standardized analysis and judgment are completed based on the clear magnitude correspondence between the current net load value and the upper limit of allowable feed into the grid connection point, and finally the grid connection point value range status of the net load change trend is obtained.

[0043] Extract the measured net load values ​​corresponding to multiple consecutive fixed collection time points in the net load change trend, statistically analyze the change pattern of this set of consecutive values, accurately distinguish the three fixed change forms of continuous increase, continuous decrease, or constant without fluctuation, and directly determine the trend characteristics of net load change based on the change form.

[0044] The grid connection point value range status and trend characteristics of net load change are combined and matched according to the fixed corresponding rules pre-set in the photovoltaic energy storage power station dispatch system. The only suitable dispatch mode for the photovoltaic energy storage power station is determined from the three standard modes of peak shaving mode, valley filling mode and maintenance mode.

[0045] When the photovoltaic energy storage power station is in peak shaving mode, it strictly follows the fixed order of priority of distributed energy storage nodes in the dynamic priority queue from high to low, selects the corresponding node in turn, and issues a discharge start command to execute the discharge operation, and continues to maintain the discharge operation state until the current net load value is stable within the safe value range below the upper limit reference value of the grid connection point.

[0046] When the photovoltaic energy storage power station is in valley filling mode, it strictly follows the fixed arrangement order of distributed energy storage nodes in the dynamic priority queue from low to high priority, selects the corresponding node in turn, and issues a charging start command to execute the charging operation, and continues to maintain the charging operation until the current net load value is stable within the safe value range above the lower limit reference value allowed by the grid connection point.

[0047] Itm4. During the scheduling process, the dynamic priority queue is adaptively corrected based on the real-time charge status and health status of the distributed energy storage nodes, and the corrected dynamic priority queue is then used to optimize the subsequent scheduling order. In this embodiment of the invention, the step of adaptively correcting the dynamic priority queue based on the real-time state of charge and health status of the distributed energy storage nodes during the scheduling process, and then using the corrected dynamic priority queue to continuously optimize the subsequent scheduling order, includes: After the scheduling step of the photovoltaic energy storage power station is completed, the real-time state of charge is updated by superimposing the real-time state of charge change of the distributed energy storage nodes. Using the health status of distributed energy storage nodes as fixed attribute values, the updated state of charge is re-executed with state membership determination and fuzzy reasoning to obtain the corrected priority label of the updated state of charge. Based on the priority label, the dynamic priority queue is adaptively corrected.

[0048] The rolling optimization of the subsequent scheduling order includes: The modified priority label will be used as the basis for scheduling distributed energy storage nodes in the next scheduling step. The distributed energy storage nodes are reordered according to the priority values ​​in the corrected priority labels, and the corrected dynamic priority queue is used as the input for the next scheduling step. The complete cycle of scheduling mode determination, charging and discharging operation execution, state of charge acquisition, and priority queue correction is repeated until the current scheduling cycle of the photovoltaic energy storage power station ends.

[0049] After all the charging and discharging operations corresponding to a single pre-set scheduling step of the photovoltaic energy storage power station have been completed, the real-time state of charge change generated by the node during the charging and discharging process of this scheduling step is accurately collected by the state acquisition module built into the distributed energy storage node. This change is then directly superimposed with the original real-time state of charge value collected at the previous scheduling moment of the node, completing the full and accurate update of the real-time state of charge of the distributed energy storage node, ensuring that the updated value truly reflects the current energy storage status of the node.

[0050] The health status value of the distributed energy storage node is kept constant and used as a fixed attribute value that is valid for a long time. Based on the charging and discharging threshold range preset by the photovoltaic energy storage power station scheduling system, the real-time charge status value after the update is accurately determined again. The obtained charge status label is matched with all the judgment rules in the fuzzy inference rule base one by one to finally obtain the exclusive correction priority label corresponding to the updated charge status of the distributed energy storage node.

[0051] Based on the scheduling priority level represented by the correction priority tags of all distributed energy storage nodes within the photovoltaic energy storage power station, all distributed energy storage nodes are rearranged in an orderly manner according to a unified standard from high to low priority level. This automatically completes the adaptive correction of the dynamic priority queue of the photovoltaic energy storage power station, ensuring that the queue status matches the latest real-time operating status of the nodes.

[0052] The correction priority tag corresponding to each distributed energy storage node is directly set as the sole criterion for determining whether the node will participate in the charging and discharging scheduling operation when entering the next scheduling step. This ensures that the execution of all scheduling actions in the next scheduling step is based entirely on the correction priority tag as the core execution standard, without introducing any other additional judgment conditions.

[0053] Based on the precise priority values ​​corresponding to the priority tags of all distributed energy storage nodes, all nodes are uniformly rearranged using a fixed sorting rule from high to low priority values. The resulting corrected dynamic priority queue is directly used as the initial input for the next scheduling step of the photovoltaic energy storage power station, providing core data support for subsequent scheduling processes.

[0054] The entire scheduling process is continuously executed in sequence according to a fixed procedure, including determining the scheduling mode, executing the corresponding charging and discharging operation, collecting the real-time charge status of nodes, and correcting the dynamic priority queue. This complete scheduling process is executed continuously and repeatedly as an independent loop unit until the total duration of the current scheduling cycle preset by the photovoltaic energy storage power station is completely exhausted, at which point the entire loop operation process is automatically terminated.

[0055] Itm5. During the current scheduling cycle of the photovoltaic energy storage power station, when the state of charge of any distributed energy storage node reaches the preset charging and discharging threshold, the distributed energy storage node will be temporarily removed from the dynamic priority queue, and the balancing protection action of the distributed energy storage node will be triggered.

[0056] In this embodiment of the invention, the step of, within the current scheduling cycle of the photovoltaic energy storage power station, when the state of charge of any distributed energy storage node reaches a preset charging and discharging threshold, includes: The real-time state of charge (SOC) values ​​of distributed energy storage nodes in the dynamic priority queue are traversed at fixed time intervals. The real-time state of charge value is compared with the preset charge and discharge threshold in real time to obtain the state of charge range of the real-time state of charge value.

[0057] The step of temporarily removing the distributed energy storage node from the dynamic priority queue and triggering the distributed energy storage node's load balancing protection action includes: When the current state of charge value read is within the normal range of the state of charge, the original position in the dynamic priority queue remains unchanged. When the current state of charge value read deviates from the normal range of the state of charge, it is recorded as an abnormal state node. Store the node identifier and current state of charge value of the abnormal node in the temporary monitoring record table, and suspend the participation qualification of the abnormal node in subsequent scheduling instructions until the state of charge of the abnormal node returns to the normal state of charge range.

[0058] According to the fixed monitoring time intervals pre-set by the photovoltaic energy storage power station dispatch system before operation based on the energy storage node safety monitoring specifications and dispatch response requirements, and strictly following the existing arrangement order of the dynamic priority queue, each distributed energy storage node in the queue is accessed one by one. The current real-time state of charge value of the node is synchronously collected through the dedicated power acquisition module configured in the distributed energy storage node, and the value is stored in real time in the real-time status recording unit of the power station dispatch system to complete the complete recording.

[0059] A fixed monitoring interval is pre-set within the photovoltaic energy storage power station dispatch system. This interval is determined based on the safety monitoring frequency standards for distributed energy storage batteries and the real-time response requirements of grid-connected dispatch. The dispatch system strictly follows this predetermined monitoring interval and retrieves the state of charge (SOC) values ​​uploaded in real time by the built-in power acquisition module of each distributed energy storage node in the queue according to the predetermined node arrangement order of the dynamic priority queue. This completes the full coverage and verification of the SOC values ​​of all distributed energy storage nodes in the queue.

[0060] Photovoltaic energy storage power stations pre-set fixed charging upper and discharging lower thresholds based on the safety operation specifications of energy storage batteries and the safety standards for grid connection. These serve as unified benchmarks for determining charging and discharging. The dispatching system compares the real-time state of charge (SOC) value of each distributed energy storage node with the predetermined charging upper and discharging lower thresholds one by one. Based on three criteria—the value falling between the charging upper and discharging lower thresholds, the value being equal to or exceeding the charging upper threshold, and the value being equal to or below the discharging lower threshold—the system accurately classifies the SOC interval category to which the corresponding real-time SOC value belongs.

[0061] The intermediate value range defined by the upper limit of charging and the lower limit of discharging is used as a fixed criterion for determining the normal state of charge range. When the scheduling system determines that the real-time state of charge value of the distributed energy storage node falls completely within this fixed value range, it does not change the existing position of the distributed energy storage node in the dynamic priority queue, nor does it adjust the arrangement order of other nodes in the dynamic priority queue, thus maintaining the original structure of the queue and the node arrangement relationship as constant.

[0062] Using the boundary of the normal state of charge interval formed by the upper limit of charging threshold and the lower limit of discharging threshold as the criterion, when the real-time state of charge value of a distributed energy storage node is higher than the upper limit of charging threshold or lower than the lower limit of discharging threshold, the distributed energy storage node is uniformly classified into a special abnormal state node category, and the special marking and classification operation of the node's operating status is completed.

[0063] The dispatch system retrieves the unique identification number of each node designated as having an abnormal state as its node identifier. Simultaneously, it captures the raw value of the node's current real-time state of charge. Both types of information are entered into a temporary monitoring record table specially built by the photovoltaic energy storage power station to complete information archiving and retention. At the same time, the abnormal node is restricted from receiving and executing all subsequent peak shaving and valley filling related dispatch instructions. This restriction is maintained until the node's state of charge value is detected to have fallen back to the normal state of charge range between the upper charging threshold and the lower discharging threshold during subsequent monitoring. Only then is the corresponding dispatch participation qualification restriction lifted.

[0064] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0065] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection, characterized in that, The method includes: Itm1. Based on the historical status records of photovoltaic energy storage power stations, the net load power of photovoltaic energy storage power stations is extrapolated over time to obtain the net load change trend of photovoltaic energy storage power stations; Itm2. Determine the state membership of the photovoltaic energy storage power station based on its current state of charge and health status values ​​to obtain the dynamic priority queue of the photovoltaic energy storage power station; Itm3. Based on the net load change trend, determine the dispatch mode of the photovoltaic energy storage power station and execute the corresponding dispatch process of the dispatch mode; Itm4. During the scheduling process, the dynamic priority queue is adaptively corrected based on the real-time charge status and health status of the distributed energy storage nodes, and the corrected dynamic priority queue is then used to optimize the subsequent scheduling order. Itm5. During the current scheduling cycle of the photovoltaic energy storage power station, when the state of charge of any distributed energy storage node reaches the preset charging and discharging threshold, the distributed energy storage node will be temporarily removed from the dynamic priority queue, and the balancing protection action of the distributed energy storage node will be triggered.

2. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 1, characterized in that, The process involves using historical records of photovoltaic energy storage power stations to perform time-series extrapolation of the net load power of these stations, resulting in the net load change trend. This includes: Read the historical net load sequence from the historical operation database of the photovoltaic energy storage power station; By performing a collaborative analysis of the net load feature vector in the historical net load sequence and the current net load sequence of the photovoltaic energy storage power station, the net load change trend of the photovoltaic energy storage power station can be obtained.

3. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 1, characterized in that, The process of determining the state membership of the photovoltaic energy storage power station based on its current state of charge and health status to obtain a dynamic priority queue for the photovoltaic energy storage power station includes: Based on the charging and discharging threshold range of the photovoltaic energy storage power station, determine the charge state affiliation label and health state affiliation label of the distributed energy storage nodes in the photovoltaic energy storage power station. Combine the charge state affiliation tag and the health state affiliation tag into a two-dimensional affiliation tag pair; Fuzzy reasoning is performed on the two-dimensional membership label pairs to obtain the dynamic priority queue of the photovoltaic energy storage power station.

4. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 3, characterized in that, The process of performing fuzzy reasoning on two-dimensional membership label pairs to obtain the dynamic priority queue of the photovoltaic energy storage power station includes: The charge state membership label and health state membership label of the distributed energy storage node are input into the fuzzy inference rule base of the photovoltaic energy storage power station for matching calculation to obtain the priority value of the two-dimensional membership label pair. A dynamic priority queue for photovoltaic energy storage power stations is constructed based on priority values.

5. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 4, characterized in that, The formula for calculating the priority value is as follows: ； In the formula, This represents the priority value of the distributed energy storage node. This represents the total number of rules in the fuzzy reasoning rule base. For the first The matching degree of the antecedent of the charge state under the rule, For the first The health status antecedent matching degree of the rule. The preset semantic distance attenuation coefficient, For the first The comprehensive semantic distance of the rules, For the first The priority weight of the consequent of a rule. It is a natural constant.

6. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 1, characterized in that, The process of determining the dispatch mode of the photovoltaic energy storage power station based on the net load change trend and executing the corresponding dispatch process for the dispatch mode includes: By comparing and analyzing the current net load value in the net load change trend with the reference value of the upper limit of allowable feed-in at the grid connection point of the photovoltaic energy storage power station, the grid connection point value range status of the net load change trend is obtained. Based on the changing pattern of net load, determine the trend characteristics of net load change. Based on the value range status and trend characteristics of the grid connection point, the dispatch mode of the photovoltaic energy storage power station is determined. The dispatch modes include peak shaving mode, valley filling mode and maintenance mode. In peak shaving mode, distributed energy storage nodes are selected sequentially from high priority to low priority according to the dynamic priority queue to perform discharge operations until the current net load value drops below the upper limit reference value allowed for grid connection. In valley filling mode, distributed energy storage nodes are selected sequentially from low priority to high priority according to the dynamic priority queue to perform charging operations until the current net load value rises above the lower limit reference value allowed by the grid connection point.

7. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 1, characterized in that, During the scheduling process, the dynamic priority queue is adaptively adjusted based on the real-time state of charge and health status of the distributed energy storage nodes. The adjusted dynamic priority queue then undergoes rolling optimization of the subsequent scheduling order, including: After the scheduling step of the photovoltaic energy storage power station is completed, the real-time state of charge is updated by superimposing the real-time state of charge change of the distributed energy storage nodes. Using the health status of distributed energy storage nodes as fixed attribute values, the updated state of charge is re-executed with state membership determination and fuzzy reasoning to obtain the corrected priority label of the updated state of charge. Based on the priority label, the dynamic priority queue is adaptively corrected.

8. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 7, characterized in that, The rolling optimization of the subsequent scheduling order includes: The modified priority label will be used as the basis for scheduling distributed energy storage nodes in the next scheduling step. The distributed energy storage nodes are reordered according to the priority values ​​in the corrected priority labels, and the corrected dynamic priority queue is used as the input for the next scheduling step. The complete cycle of scheduling mode determination, charging and discharging operation execution, state of charge acquisition, and priority queue correction is repeated until the current scheduling cycle of the photovoltaic energy storage power station ends.

9. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 1, characterized in that, During the current scheduling cycle of the photovoltaic energy storage power station, when the state of charge of any distributed energy storage node reaches a preset charging / discharging threshold, the following includes: The real-time state of charge (SOC) values ​​of distributed energy storage nodes in the dynamic priority queue are traversed at fixed time intervals. The real-time state of charge value is compared with the preset charge and discharge threshold in real time to obtain the state of charge range of the real-time state of charge value.

10. The peak-shaving and valley-filling optimization scheduling method for distributed energy storage power stations during photovoltaic grid connection as described in claim 9, characterized in that, The step of temporarily removing the distributed energy storage node from the dynamic priority queue and triggering the distributed energy storage node's load balancing protection action includes: When the current state of charge value read is within the normal range of the state of charge, the original position in the dynamic priority queue remains unchanged. When the current state of charge value read deviates from the normal range of the state of charge, it is recorded as an abnormal state node. Store the node identifier and current state of charge value of the abnormal node in the temporary monitoring record table, and suspend the participation qualification of the abnormal node in subsequent scheduling instructions until the state of charge of the abnormal node returns to the normal state of charge range.

Images