An artificial intelligence-based power energy storage dynamic scheduling system and method

Abstract

The application discloses an artificial intelligence-based power energy storage dynamic scheduling system and method, relates to the technical field of power energy storage dynamic scheduling, and comprises the following steps: constructing a power energy storage demand model of a power grid; analyzing the performance deviation condition of energy storage battery groups under different operating environments, and determining the deviation data of the energy storage battery groups; predicting the power energy storage demand of the power grid by using the power energy storage demand model, combining the deviation data, analyzing the performance deviation condition of the energy storage battery groups, analyzing the adaptability of different battery clusters in different energy storage battery groups to scheduling instructions, and determining the scheduling strategy of the energy storage battery groups in the power grid; and performing operating control on the energy storage battery groups in the power grid according to the energy storage scheduling strategy, and scheduling the power energy storage of the power grid, so that the power energy storage demand of the power grid is efficiently met, the health of the energy storage battery groups is greatly guaranteed, and the service life and safety of the energy storage battery groups are improved.

Description

Technical Field

[0001] This invention relates to the field of dynamic dispatching technology for power energy storage, specifically a dynamic dispatching system and method for power energy storage based on artificial intelligence. Background Technology

[0002] Electricity storage is essentially the process of storing electrical energy using certain technologies or devices and releasing it when needed. However, in reality, electricity demand fluctuates significantly across different seasons, weeks, and even days. Furthermore, the power generation capacity of new energy sources like wind and solar power varies significantly with weather conditions. Therefore, the power grid needs to dynamically schedule electricity storage based on actual electricity demand and power generation status. Dynamic scheduling of electricity storage in the power grid offers advantages including, but not limited to, the following: 1. Addressing real-time fluctuations in electricity supply and demand: The power grid needs to maintain a balance between power generation and consumption at all times. Dynamic scheduling of electricity storage can quickly adjust the charging and discharging state, maintaining grid stability; 2. Improving the stability of new energy power generation: Dynamic scheduling of energy storage can reduce the instability caused by renewable energy sources such as electricity and photovoltaics; 3. Enhancing grid reliability: Dynamic scheduling not only reduces grid operating costs but also ensures a reliable and sufficient supply of electricity during peak demand periods.

[0003] During grid operation, it is necessary to deal not only with changes in electricity demand but also with the instability of new energy power generation. Therefore, dynamic scheduling of power storage occurs frequently. It is not only necessary to realize the scheduling plan of power generation and energy storage within a preset time period, but also to change the power within minutes. This requires frequent scheduling operations for energy storage battery packs. However, the health status and operating conditions of different batteries in the energy storage battery pack are different, and different energy storage battery packs can also affect each other. If the operation is simply carried out according to the scheduling instructions, it will not only affect the service life of the batteries in the energy storage battery pack, but may even lead to safety problems of the batteries in the energy storage battery pack. Summary of the Invention

[0004] The purpose of this invention is to provide an artificial intelligence-based dynamic scheduling system and method for power storage, in order to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a dynamic scheduling method for power storage based on artificial intelligence, the method comprising: Step S1: Obtain historical power change records and historical operating status records of the power grid, and construct a power storage demand model for the power grid; Step S2: Obtain the historical operation records of the energy storage battery packs in the power grid, analyze the performance deviation conditions of the energy storage battery packs under different operating environments, and determine the deviation data of the energy storage battery packs; Step S3: Obtain the power grid operation data of the power grid in the current cycle, use the power energy storage demand model to predict the power energy storage demand of the power grid, obtain the energy storage demand prediction data, and combine the deviation data to analyze the performance deviation conditions of the energy storage battery packs, and analyze the adaptability of different battery clusters in different energy storage battery packs in the power grid to the scheduling instructions, determine the scheduling strategy of the energy storage battery packs in the power grid, and obtain the energy storage scheduling strategy; Step S4: According to the energy storage scheduling strategy, perform operation control on the energy storage battery packs in the power grid and schedule the power energy storage of the power grid.

[0006] Further, Step S3 includes: According to the energy storage demand prediction data, obtain the overall energy storage dispatching order data of each energy storage battery pack in the power grid, and obtain the overall power demand value P of the power grid in the current period of the current cycle total ; Obtain the deviation data of a certain energy storage battery pack in the power grid, and calculate the performance deviation value H of a certain energy storage battery pack; Set the performance deviation threshold H △ , when H △ <H, it is determined that the energy storage battery pack will not have performance deviation, and a certain energy storage battery pack is recorded as the characteristic energy storage battery pack of the power grid in the current period. Otherwise, obtain a certain marked working condition record set corresponding to the performance deviation value H; Calculate the outlier R of a certain energy storage battery pack, set the outlier threshold r, when R>r, it is determined that a certain energy storage battery pack has an abnormal risk. Otherwise, it is determined that a certain energy storage battery pack does not have an abnormal risk, and a certain energy storage battery pack is recorded as the characteristic energy storage battery pack; Obtain the energy storage battery power order set ζ, and according to the order in the energy storage battery power order set ζ, obtain the target energy storage battery pack of the power grid in the current period; Take the characteristic power value of the target energy storage battery pack in the current period as the total energy storage power of the target energy storage battery pack in the current period; Obtain the unit characteristic power and battery health score of each battery cluster of the target energy storage battery pack in the current period, obtain the battery health score G of a certain battery cluster of the target energy storage battery pack, and set the battery health score threshold G ◇ , when G<G ◇ At this time, mark a certain battery cluster, and obtain the target power of a certain battery cluster in the current period from the historical operation records of the target energy storage battery pack; When G≥G ◇In this case, a certain battery cluster is not marked, and the sum of the target power V of the marked battery clusters in the target energy storage battery pack is obtained. sum The sum of the total energy storage power and the target power is V. sum The total power Z is obtained. sum The target power of the battery clusters of the unmarked target energy storage battery pack is randomly generated by the platform. The target power of each battery cluster in each target energy storage battery pack is obtained in the current time period to obtain the battery operation data of the power grid in the current time period. The battery operation data of the power grid in each unit time period in the current cycle is obtained to obtain the energy storage scheduling strategy of the power grid in the current cycle. The above steps not only accurately obtain the power storage demand of the power grid in the current cycle, but also solve a series of problems such as which energy storage battery packs need to be dispatched to meet the power storage demand of the power grid in the current cycle, how much workload should be allocated to the dispatched energy storage battery packs (i.e., the power corresponding to charging and discharging), and how to operate and control different battery clusters within the energy storage battery packs. This makes the generation of energy storage dispatch not only meet the needs of the power grid, but also effectively ensure the safe and stable operation of the batteries.

[0007] Furthermore, step S2 includes: Obtain historical operation records from the energy storage battery pack, extract data on various operating condition parameters of the energy storage battery pack from the historical operation records, and construct the operating condition feature vector of the energy storage battery pack from the historical operation records. Data on various performance parameters of the energy storage battery pack are obtained from historical operation records, and a performance feature vector of the energy storage battery pack in the historical operation records is constructed. Calculate the approximate operating conditions between various historical operating records of the energy storage battery pack, and set an approximate operating condition threshold S. When the approximate operating condition value S between one historical operating record and another of the energy storage battery pack is... ▽ When S ▽ If S >, then it is determined that the operating conditions of a certain historical operating record are similar to those of another historical operating record; otherwise, it is determined that the operating conditions of a certain historical operating record are not similar to those of another historical operating record. The historical operating records that are similar to a certain historical operating record are obtained from each historical operating record and aggregated to obtain the set of approximate operating condition records α of a certain historical operating record. Calculate the average performance offset C' among several historical operation records in the approximate operating condition record set α, and set the performance offset threshold C. ◇ When C'>C ◇If it is determined that the performance of the energy storage battery pack has deviated in several historical operation records in the working condition approximation record set α, the working condition approximation record set α is marked and recorded as the marked working condition record set. Otherwise, it is determined that the performance of the energy storage battery pack has not deviated in several historical operation records in the determination working condition approximation record set α; Obtain and collect each marked working condition record set in which the energy storage battery pack has performance deviation in the historical operation records to obtain the deviation data of the energy storage battery pack.

[0008] Further, step S1 includes: Obtain the historical power change records and historical operation status records of the power grid in each historical period; Set the unit time length, obtain the total power generation w and total load power q of the power grid within the unit time length of the historical period from the historical power change records, and calculate the power energy storage value of the power grid within the unit time length in the historical power change records ; Obtain the data of each operation parameter of the power grid within the historical period from the historical operation status records, and construct the power energy storage demand model of the power grid according to the historical power change records and historical operation status records of the power grid in each historical period. The specific construction process of the power energy storage demand model is as follows: Collect the historical power change records and historical operation status records with the same historical period in the power grid and record them as the power grid record group within the historical period; Align the power energy storage value of the historical power change record and each operation parameter in the historical operation status record in the power grid record group according to the unified time stamp, use the power energy storage value in the historical power change record as the output data of the power energy storage model, and use each operation parameter in the historical operation status record as the input data of the power energy storage model. Obtain the power grid record groups of the power grid in each historical period, and divide the power grid record groups of the power grid in each historical period into a training set and a test set according to a preset ratio. Use the training set to train the model using the intelligent algorithm in the power energy storage demand model, use the test set to calculate the model error value A of the power energy storage demand model, set the model error threshold A´. When A < A´, it is determined that the construction of the power energy storage demand model is completed. Otherwise, adjust the model parameters in the power energy storage demand model until A < A´.

[0009] Further, step S4 includes: Obtain the energy storage dispatching strategy of the power grid in the current period, and obtain the target power of each energy storage battery pack in the power grid within each unit time length in the current period from the energy storage dispatching strategy; Based on the target power of each energy storage battery pack in the power grid within each unit of time in the current cycle, the charging and discharging of each energy storage battery pack in the power grid is controlled, and the power storage of the power grid is dynamically dispatched within the current cycle.

[0010] To better implement the above methods, an artificial intelligence-based dynamic dispatch system for power energy storage is also proposed. The system includes an energy storage demand model construction module, an offset analysis module, an energy storage dispatch strategy determination module, and a dynamic dispatch module. The energy storage demand model construction module is used to acquire historical power change records and historical operating status records of the power grid, and to construct the power energy storage demand model of the power grid. The offset analysis module is used to acquire historical operating records of energy storage battery packs in the power grid, analyze the performance offset of energy storage battery packs under different operating environments, and determine the offset data of energy storage battery packs. The energy storage dispatch strategy determination module is used to predict the power storage demand of the power grid using the power storage demand model, and analyze the performance deviation of the energy storage battery packs in combination with offset data, and analyze the adaptability of different battery clusters in different energy storage battery packs in the power grid to dispatch commands, so as to obtain the energy storage dispatch strategy. The dynamic scheduling module is used to control the operating status of different energy storage battery packs in the power grid according to the energy storage scheduling strategy, and to dynamically schedule the power grid's energy storage.

[0011] Furthermore, the energy storage demand model construction module includes a behavior record acquisition unit and an energy storage demand model construction unit; The behavior record acquisition unit is used to acquire historical power change records and historical operating status records of the power grid; The energy storage demand model building unit is used to build the power grid's energy storage demand model based on the grid's historical power change records and historical operating status records.

[0012] Furthermore, the offset analysis module includes a vector construction unit and an offset analysis unit; The vector construction unit is used to acquire historical operating records of the energy storage battery pack and construct the operating condition feature vector and performance feature vector of the energy storage battery pack in the historical operating records. The offset analysis unit is used to analyze the performance offset of the energy storage battery pack under different operating environments based on the operating condition feature vector and the performance feature vector, and to determine the offset data of the energy storage battery pack.

[0013] Furthermore, the energy storage scheduling strategy determination module includes a scheduling adaptation analysis unit and an energy storage scheduling strategy determination unit; The scheduling adaptation analysis unit is used to acquire grid operation data in the current cycle, predict the grid's energy storage demand using the power storage demand model, obtain energy storage demand prediction data, and analyze the performance deviation of energy storage battery packs in combination with offset data, and analyze the adaptability of different battery clusters in different energy storage battery packs in the grid to scheduling commands. The energy storage dispatch strategy determination unit is used to determine the dispatch strategy of energy storage battery packs in the power grid, and obtain the energy storage dispatch strategy.

[0014] Furthermore, the dynamic scheduling module includes a dynamic scheduling unit; The dynamic scheduling unit is used to acquire the energy storage scheduling strategy of the power grid in the current cycle, and control the charging and discharging of each energy storage battery pack in the power grid according to the energy storage scheduling strategy, so as to dynamically schedule the power energy storage of the power grid in the current cycle.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: By recording historical power changes and historical operating status in the power grid, a power storage demand model is constructed, and the performance deviation of the energy storage battery pack under different operating environments is determined. In the current cycle, not only is the power storage demand of the power grid predicted through the power storage demand model, but also the adaptability of different battery clusters in different energy storage battery packs in the power grid to dispatch instructions is analyzed. This determines which battery clusters in the power storage battery packs of the power grid need to perform charging and discharging operations at what power in specific energy storage dispatch instructions. Thus, dynamic dispatch of power storage is truly realized, which not only efficiently meets the energy storage demand of the power grid, but also greatly protects the health of the energy storage battery packs and improves their service life and safety. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the scheduling strategy determination process of an artificial intelligence-based dynamic scheduling method for power storage according to the present invention. Figure 2 This is a schematic diagram of a dynamic scheduling system for power storage based on artificial intelligence, according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example: Figures 1-2 As shown, the present invention provides a technical solution, a dynamic scheduling method for power storage based on artificial intelligence, the method comprising: Step S1: Obtain the historical power change records and historical operation status records of the power grid, and construct a power energy storage demand model for the power grid; Among them, step S1 includes: Obtain the historical power change records and historical operation status records of the power grid in each historical period; Set a unit time period, obtain the total power generation w and the total load power q of the power grid within the unit time period of the historical period from the historical power change records, and calculate the power energy storage value of the power grid within the unit time period in the historical power change records ; For example, the total load power is: The integral of the electric power taken by each power consumption unit in the power grid from the power system within the unit time period at a certain time period, where the power consumption units include industrial parks, residential houses, etc.; For example, if the value of the power energy storage value is positive, it is necessary to store power with a battery energy storage bank, and if it is negative, it is necessary for the battery energy storage bank to discharge; Obtain the data of various operation parameters of the power grid in the historical period from the historical operation status records, and construct a power energy storage demand model for the power grid according to the historical power change records and historical operation status records of the power grid in each historical period. The specific construction process of the power energy storage demand model is as follows: For example, the various operation parameters include weather temperature, date type (working day, holiday), electricity price signal, etc.; Collect the historical power change records and historical operation status records with the same historical period in the power grid, and record them as the power grid record group within the historical period; Align the power energy storage value of the historical power change record and the various operation parameters in the historical operation status record in the power grid record group according to the unified time stamp, and use the power energy storage value in the historical power change record as the output data of the power energy storage model, and use the various operation parameters in the historical operation status record as the input data of the power energy storage model. Obtain the power grid record groups of the power grid in each historical period, and divide the power grid record groups of the power grid in each historical period into a training set and a test set according to a preset ratio. Use the training set to train the model using the intelligent algorithm in the power energy storage demand model, use the test set to calculate the model error value A of the power energy storage demand model, set the model error threshold A´. When A < A´, it is determined that the construction of the power energy storage demand model is completed. Otherwise, adjust the model parameters in the power energy storage demand model until A < A´; For example, the specific acquisition process of the model error value A is as follows: Set a unit time period, and obtain the power energy storage value within each unit time period of the historical power change record of the power grid record group in the test set; Calculate the prediction deviation value 'a' for the power grid record set in the test set: , Among them, Y i y represents the actual value of the energy storage within the i-th unit of time in the historical power change record of the power grid record group; i is the predicted value of power storage within the i-th unit of time in the historical power change records of the power grid record group, as provided by the power storage demand model; n is the total number of units of time in the historical power change records of the power grid record group. The average prediction deviation value of each power grid record group in the test set is calculated to obtain the model error value A of the power storage demand model.

[0019] Step S2: Obtain historical operation records of energy storage battery packs in the power grid, analyze the performance deviation of energy storage battery packs under different operating environments, and determine the deviation data of energy storage battery packs; Step S2 includes: Obtain historical operation records from the energy storage battery pack, extract data on various operating condition parameters of the energy storage battery pack from the historical operation records, and construct the operating condition feature vector of the energy storage battery pack from the historical operation records. Data on various performance parameters of the energy storage battery pack are obtained from historical operation records, and a performance feature vector of the energy storage battery pack in the historical operation records is constructed. For example, the specific process for constructing the operating condition feature vector of the energy storage battery pack from historical operation records is as follows: Data on various operating parameters of the energy storage battery pack are obtained from historical operation records. These parameters include the charge / discharge rate, average voltage, and average temperature of the energy storage battery. Obtain the maximum and minimum values ​​of various operating condition parameters from each historical operating record in the energy storage battery pack, and normalize the various operating condition parameters in the energy storage battery pack. According to a preset order, the various operating condition parameters of the energy storage battery pack obtained from the historical operation records are arranged to obtain the operating condition feature vector of the energy storage battery pack in the historical operation records. For example, the specific process for constructing the performance feature vector of the energy storage battery pack from historical operation records is as follows: Data on various performance parameters of the energy storage battery pack are obtained from historical operation records. These performance parameters include the DC internal resistance of the energy storage battery and the battery charging and discharging efficiency. Obtain the maximum and minimum values ​​of various performance parameters from each historical operation record in the energy storage battery pack, and normalize the various performance parameters in the energy storage battery pack. According to a preset order, the performance parameters of the energy storage battery pack obtained from the historical operation records are arranged to obtain the performance feature vector of the energy storage battery pack in the historical operation records. Calculate the approximate operating conditions between various historical operating records of the energy storage battery pack, and set an approximate operating condition threshold S. When the approximate operating condition value S between one historical operating record and another of the energy storage battery pack is... ▽ When S ▽ If S >, then it is determined that the operating conditions of a certain historical operating record are similar to those of another historical operating record; otherwise, it is determined that the operating conditions of a certain historical operating record are not similar to those of another historical operating record. The historical operating records that are similar to a certain historical operating record are obtained from each historical operating record and aggregated to obtain the set of approximate operating condition records α of a certain historical operating record. For example, the approximate operating condition value S between one historical operating record and another historical operating record. ▽ The specific calculation formula is as follows: , Where D is the operating condition feature vector of the energy storage battery pack in a certain historical operation record; D´ is the operating condition feature vector of the energy storage battery pack in another historical operation record; Calculate the average performance offset C' among several historical operation records in the approximate operating condition record set α, and set the performance offset threshold C. ◇ When C'>C ◇ If the performance of the energy storage battery pack deviates in several historical operation records in the approximate operating condition record set α, then the approximate operating condition record set α is marked and denoted as the marked operating condition record set; otherwise, it is determined that the performance of the energy storage battery pack in several historical operation records in the approximate operating condition record set α has not deviated. For example, the performance offset value C between the γth historical operation record and the βth historical operation record in the approximate operating condition record set α. (γ,β) : , Among them, E γ E represents the performance feature vector of the energy storage battery pack in the γth historical operation record. β Let be the performance feature vector of the energy storage battery pack in the βth historical operation record; The system acquires and aggregates the records of each marked operating condition in the historical operation records where the energy storage battery pack experienced performance deviations, thus obtaining the deviation data of the energy storage battery pack.

[0020] Step S3: Obtain the grid operation data in the current cycle, use the power storage demand model to predict the power storage demand of the grid, obtain the energy storage demand prediction data, and combine it with the offset data to analyze the performance offset of the energy storage battery pack, and analyze the adaptability of different battery clusters in different energy storage battery packs in the grid to the dispatch instructions, determine the dispatch strategy of the energy storage battery pack in the grid, and obtain the energy storage dispatch strategy. Step S3 includes: Based on energy storage demand forecast data, obtain the overall energy storage dispatch data for each energy storage battery pack in the power grid, and obtain the overall power demand value P of the power grid in the current period of the current cycle. total ; For example, the overall power demand value P total The specific acquisition process is as follows: Obtain grid operation data for the current cycle, input the grid operation data into the power storage demand model, and obtain energy storage demand forecast data; For example, power grid operation data includes data on various operation parameters. When the operation parameters are parameters such as weather temperature and ambient wind speed, the values ​​of weather temperature and ambient wind speed are obtained from the weather forecast of the power grid area in the current period. For example, energy storage demand forecast data includes the value of electricity storage within each unit of time in the current period; Based on energy storage demand forecast data, obtain the overall energy storage dispatch data for each energy storage battery pack in the power grid, and extract the overall power demand value P of the power grid in the current period of the current cycle from the overall energy storage dispatch data. total ; For example, the current time period specifically refers to the duration of the first unit of time within the current cycle; For example, based on energy storage demand forecast data, the overall energy storage dispatch data of each energy storage battery pack in the power grid is obtained. The specific process is as follows: The power storage value is obtained from the energy storage demand forecast data for each unit of time in the current period. When the power storage value is positive, the absolute value of the corresponding value is the total amount of electricity that each energy storage battery pack needs to store. When the power storage value is negative, the absolute value of the corresponding value is the total amount of electricity that each energy storage battery pack needs to transmit. Based on the power storage value of the power grid in each unit of time within the current cycle, the charging power and discharging power of each energy storage battery pack in each unit of time within the current cycle are obtained through the ACG platform of the power grid, and then aggregated into the overall energy storage dispatch data of each energy storage battery pack in the power grid. For example, obtain the total power required for charging or generating electricity by the power grid during the current period of the current cycle from the overall energy storage dispatching order data, and record it as the overall power demand value; Obtain the offset data of a certain energy storage battery pack in the power grid, and calculate the performance offset value H of a certain energy storage battery pack; For example, the specific calculation process of the performance offset value H is as follows: Obtain the average power of the energy storage battery pack in the historical operation records of each marked working condition record set from the offset data, and obtain several marked working condition record sets with the same power direction as the characteristic power value of a certain energy storage battery pack from the offset data, and calculate the performance offset value H of a certain energy storage battery pack: , where, λ represents the characteristic power value of a certain energy storage battery pack; λ i represents the average power of the energy storage battery pack in each historical operation record of the i-th marked working condition record set of several marked working condition record sets; n represents the total number of several marked working condition record sets with the same power direction as the characteristic power value of a certain energy storage battery pack in the offset data; Set the performance offset threshold H △ , when H △ < H, it is determined that the energy storage battery pack will not have performance offset, and a certain energy storage battery pack is recorded as the characteristic energy storage battery pack of the power grid during the current period. Otherwise, obtain a marked working condition record set corresponding to the performance offset value H; Calculate the outlier value R of a certain energy storage battery pack, set the outlier threshold r, when R > r, it is determined that a certain energy storage battery pack has an abnormal risk. Otherwise, it is determined that a certain energy storage battery pack does not have an abnormal risk, and a certain energy storage battery pack is recorded as the characteristic energy storage battery pack; For example, the specific process of calculating the outlier value R of a certain energy storage battery pack is as follows: Obtain the historical operation records of equipment abnormalities of the energy storage battery pack in a certain marked working condition record set, and calculate the outlier value of a certain energy storage battery pack , where, U is the total number of historical operation records of a certain energy storage battery pack with abnormalities in a certain marked working condition record set, U sum represents the total number of historical operation records in a certain marked working condition record set; Obtain the energy storage battery power order set ζ, and according to the order in the energy storage battery power order set ζ, obtain the target energy storage battery pack of the power grid during the current period; For example, to obtain the target energy storage battery pack of the power grid during the current period, the specific obtaining process is as follows: Successively add the characteristic power values of the characteristic energy storage battery packs to obtain the sum of the characteristic power values corresponding to each order in the energy storage battery power order set ζ, and obtain the sum of the characteristic power values greater than the overall power demand value Ptotal The minimum value of the order B min And in the energy storage battery power order set ζ, the first B min A characteristic energy storage battery pack is denoted as the target energy storage battery pack; For example, the specific process of obtaining the power order set ζ of energy storage batteries is as follows: Obtain the overall power demand value P total In terms of power direction, obtain the overall power demand value P. total The characteristic power values ​​of each characteristic energy storage battery pack in the same power direction are obtained, and the characteristic power values ​​are sorted to obtain the energy storage battery power order set ζ; For example, the overall power demand value P total Power direction: This refers to the charging and discharging behavior corresponding to the overall power demand value, when the overall power demand value P total When the power direction is positive, the power grid needs each energy storage battery pack to perform charging behavior, and the corresponding energy storage battery pack is the charging power. When the overall power demand value P total When the power direction is negative, the power grid needs each energy storage battery pack to discharge, and the corresponding discharge power of the energy storage battery pack. For example, the characteristic power value is: When the power direction of the energy storage battery pack is positive, the characteristic power value of the energy storage battery pack is the maximum charging power multiplied by a preset percentage, which gives the characteristic power value of the energy storage battery pack when the power direction is positive. When the power direction of the energy storage battery pack is negative, the characteristic power value of the energy storage battery pack is the maximum discharge power multiplied by a preset percentage, thus obtaining the characteristic power value of the energy storage battery pack when the power direction is negative. The characteristic power value of the target energy storage battery pack in the current time period is taken as the total energy storage power of the target energy storage battery pack in the current time period; Obtain the unit characteristic power and battery health score of each battery cluster of the target energy storage battery pack within the current time period, obtain the battery health score G of a certain battery cluster of the target energy storage battery pack, and set the battery health score threshold G. ◇ When G <G ◇ When a certain battery cluster is marked, the target power of that battery cluster in the current time period is obtained from the historical operation record of the target energy storage battery pack. For example, the characteristic power of a unit is: When the total energy storage power of the target energy storage battery pack is positive, the unit characteristic power of the target energy storage battery pack is the maximum charging power multiplied by a preset percentage, thus obtaining the unit characteristic power of the target energy storage battery pack when the power direction is positive. When the total energy storage power of the target energy storage battery pack is negative, the unit characteristic power of the target energy storage battery pack is the maximum discharge power multiplied by a preset percentage, thus obtaining the unit characteristic power of the target energy storage battery pack when the power direction is negative. For example, a battery cluster is a unit in an energy storage battery pack that can be managed independently, consisting of multiple batteries connected in series; For example, to obtain the target power of a specific battery cluster in the current time period from the historical operation records of the target energy storage battery pack, the specific process is as follows: The system obtains the historical unit time periods in which all performance parameters of a certain battery cluster are within the preset range from the historical operation records of the target energy storage battery pack. It also obtains the average value of the battery power of a certain battery cluster in the same direction as the total energy storage power in the current time period from each historical unit time period, and records it as the target power of a certain battery cluster in the current time period. When G≥G ◇ In this case, a certain battery cluster is not marked, and the sum of the target power V of the marked battery clusters in the target energy storage battery pack is obtained. sum The sum of the total energy storage power and the target power is V. sum The total power Z is obtained. sum The target power of the battery clusters of the unmarked target energy storage battery pack is randomly generated by the platform. Among them, the sum of the target power of the battery clusters randomly generated by the platform as unmarked target energy storage battery packs equals the total power Z. sum Furthermore, the target power of the battery clusters randomly generated by the platform as unmarked target energy storage battery packs is less than or equal to the unit characteristic power of the battery cluster; For example, the specific characteristic power of a battery cluster is: When the total energy storage power of the target energy storage battery pack is positive, the unit characteristic power of the target energy storage battery pack is the maximum charging power multiplied by a preset percentage, thus obtaining the unit characteristic power of the target energy storage battery pack when the power direction is positive. When the total energy storage power of the target energy storage battery pack is negative, the unit characteristic power of the target energy storage battery pack is the maximum discharge power multiplied by a preset percentage, thus obtaining the unit characteristic power of the target energy storage battery pack when the power direction is negative. For example, the specific formula for calculating the battery health score G of a single cell is:

[0021] L represents the actual capacity of the battery cluster during the current cycle; L △ This refers to the factory rated capacity of the battery cluster; The target power of each battery cluster in each target energy storage battery pack is obtained to obtain the battery operation data of the power grid in the current time period. The battery operation data of the power grid in each unit time period in the current cycle is obtained to obtain the energy storage scheduling strategy of the power grid in the current cycle.

[0022] Step S4: Based on the energy storage dispatch strategy, control the operation of the energy storage battery packs in the power grid and dispatch the power storage of the power grid.

[0023] Step S4 includes: Obtain the grid's energy storage dispatch strategy in the current cycle, and obtain the target power of each energy storage battery pack in the grid in each unit of time in the current cycle from the energy storage dispatch strategy; Based on the target power of each energy storage battery pack in the power grid within each unit of time in the current cycle, the charging and discharging of each energy storage battery pack in the power grid is controlled, and the power storage of the power grid is dynamically dispatched within the current cycle.

[0024] To better implement the above methods, an artificial intelligence-based dynamic dispatch system for power energy storage is also proposed. The system includes an energy storage demand model construction module, an offset analysis module, an energy storage dispatch strategy determination module, and a dynamic dispatch module. The energy storage demand model construction module is used to acquire historical power change records and historical operating status records of the power grid, and to construct the power energy storage demand model of the power grid. The offset analysis module is used to acquire historical operating records of energy storage battery packs in the power grid, analyze the performance offset of energy storage battery packs under different operating environments, and determine the offset data of energy storage battery packs. The energy storage dispatch strategy determination module is used to predict the power storage demand of the power grid using the power storage demand model, and analyze the performance deviation of the energy storage battery packs in combination with offset data, and analyze the adaptability of different battery clusters in different energy storage battery packs in the power grid to dispatch commands, so as to obtain the energy storage dispatch strategy. The dynamic scheduling module is used to control the operating status of different energy storage battery packs in the power grid according to the energy storage scheduling strategy, and to dynamically schedule the power grid's energy storage.

[0025] The energy storage demand model construction module includes a behavior record acquisition unit and an energy storage demand model construction unit. The behavior record acquisition unit is used to acquire historical power change records and historical operating status records of the power grid; The energy storage demand model building unit is used to build the power grid's energy storage demand model based on the grid's historical power change records and historical operating status records.

[0026] The offset analysis module includes a vector construction unit and an offset analysis unit. The vector construction unit is used to acquire historical operating records of the energy storage battery pack and construct the operating condition feature vector and performance feature vector of the energy storage battery pack in the historical operating records. The offset analysis unit is used to analyze the performance offset of the energy storage battery pack under different operating environments based on the operating condition feature vector and the performance feature vector, and to determine the offset data of the energy storage battery pack.

[0027] The energy storage scheduling strategy determination module includes a scheduling adaptation analysis unit and an energy storage scheduling strategy determination unit. The scheduling adaptation analysis unit is used to acquire grid operation data in the current cycle, predict the grid's energy storage demand using the power storage demand model, obtain energy storage demand prediction data, and analyze the performance deviation of energy storage battery packs in combination with offset data, and analyze the adaptability of different battery clusters in different energy storage battery packs in the grid to scheduling commands. The energy storage dispatch strategy determination unit is used to determine the dispatch strategy of energy storage battery packs in the power grid, and obtain the energy storage dispatch strategy.

[0028] The dynamic scheduling module includes a dynamic scheduling unit; The dynamic scheduling unit is used to acquire the energy storage scheduling strategy of the power grid in the current cycle, and control the charging and discharging of each energy storage battery pack in the power grid according to the energy storage scheduling strategy, so as to dynamically schedule the power energy storage of the power grid in the current cycle.

[0029] 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 invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Images