Integrated power supply system and scheduling method thereof

Abstract

The invention discloses an integrated power supply system and a scheduling method thereof, and belongs to the technical field of power supply systems, and the method comprises the steps: collecting the operation state data of a micro-grid in real time; establishing a storage battery stress factor table, identifying an effective charge-discharge cycle, calculating a life loss value, and converting the life loss value into life loss cost; the method comprises the following steps of: constructing an optimization model containing life loss cost by taking minimization of total operation cost as a target, and generating an optimal scheduling instruction set in each scheduling period by adopting a model prediction control framework and an improved particle swarm optimization algorithm. According to the invention, through real-time quantitative evaluation and optimal scheduling of the service life of the storage battery, the economy and reliability of the whole life cycle of the system are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power supply systems, and particularly relates to an integrated power supply system and a scheduling method thereof. BACKGROUND

[0002] In order to continuously supply power to key equipment and save energy, a hybrid micro-grid mainly composed of solar energy and wind energy and supplemented by a diesel generator is often constructed. In this system, the intelligent integrated power supply as a core control and energy scheduling unit of the micro-grid is an important part of maintaining stable power supply of the power grid.

[0003] At present, the existing power supply equipment energy management method mainly focuses on real-time balance of power and short-term operation economy (such as minimization of diesel consumption), but generally ignores fine management of the integrated power supply. Due to the randomness of renewable energy input and the volatility of load demand, the storage battery is in a non-standard random charging and discharging condition for a long time. Compared with regular standard cycles, this random condition will significantly cause instability of the integrated power supply, affecting the safe and stable operation of the power grid.

[0004] However, the existing technology lacks effective models and means to quantitatively determine the "instantaneous loss" of the battery life caused by a specific charging and discharging operation in real time. Although there are offline evaluation methods for the state of health of the battery, they cannot meet the real-time requirements of online scheduling. More importantly, the existing technology does not regard "life loss" as a key optimization variable and real-time participate in the decision-making of energy scheduling. This may cause the system operation strategy to frequently perform damaging shallow charging and discharging on the storage battery for short-term saving of a small amount of diesel, and from the perspective of the whole life cycle, the heavy cost of rapid decline of the battery life is paid, and the economy is even worse. SUMMARY

[0005] To solve the above problems, the application provides an integrated power supply system scheduling method to solve the problems in the background art.

[0006] In order to achieve the above application purpose, in a first aspect, the application provides an integrated power supply system scheduling method, comprising: Step S1, real-time collection of micro-grid operation state data, the operation state data including photovoltaic power generation power, wind power generation power, load demand power, diesel generator state, operation state, state of charge, temperature, voltage and current of the power supply equipment storage battery; Step S2, establishment of a stress factor table of the storage battery, identification and analysis of effective charging and discharging cycles of the storage battery within a preset time window based on current data of the storage battery, calculation of a loss value within the time period of the effective charging and discharging cycle based on the stress factor table, conversion of the loss value into a battery life loss cost, and updating of the stress factor table every preset time interval. Step S3, an optimization model containing life loss cost is established with the total operation cost as the objective function, the total operation cost including diesel generator fuel cost, battery life loss cost, power purchase cost from the main power grid and renewable energy abandonment penalty cost, an improved particle swarm optimization algorithm is used to solve the optimization model to obtain the optimal scheduling instruction set based on the real-time operation state data and short-term prediction information of the microgrid in each optimization period.

[0007] Further, the loss value in the time period of the effective charge-discharge cycle includes the following steps: Performing extreme value analysis on the current time series data of the battery, identifying all local maximum points and local minimum points to form a peak-valley feature point sequence, identifying complete cycles based on the peak-valley feature point sequence, and calculating the depth value of the complete cycle, matching the depth value of the identified complete cycle with the pre-established battery stress factor table to obtain the loss weight value corresponding to each depth range value, and accumulating the loss weight value of each complete cycle to obtain the loss value in the time period analyzed.

[0008] Further, identifying complete cycles based on the peak-valley feature point sequence includes the following steps: Selecting four consecutive feature points from the peak-valley feature point sequence at a time, calculating the first difference between the first feature point and the second feature point, the second difference between the second feature point and the third feature point, and the third difference between the third feature point and the fourth feature point, when the second difference is less than or equal to the first difference and less than or equal to the third difference, determining that the second feature point to the third feature point constitutes a complete cycle, taking the second difference as the depth value of the complete cycle, removing the feature points constituting the complete cycle from the peak-valley feature point sequence, and continuing to identify complete cycles from the removed peak-valley feature point sequence until no new complete cycle can be identified.

[0009] Further, updating the stress factor table includes the following steps: Establishing a battery loss prediction model, the prediction model taking the historical operation state data of the battery as input features and outputting the loss prediction value of the battery in a future period of time; Based on the prediction model, the loss prediction value of the future period of time is obtained, the calculated loss value of the corresponding period of time is obtained, the calculated loss value and the loss prediction value are compared to obtain the deviation value, and if the deviation value exceeds a preset threshold, the stress factor table is adjusted based on the deviation value and the historical working conditions of the battery.

[0010] Further, establishing a multi-objective optimization model containing life loss cost includes the following steps: Define decision variables, the decision variables include diesel engine output power, battery charging and discharging power, renewable energy scheduling ratio, load reduction, set the objective function as minimizing the total operation cost, set the constraint conditions, the constraint conditions include power balance constraint, equipment operation constraint and battery life constraint, the battery life loss cost in the total cost is the battery life loss amount multiplied by the unit loss cost of the battery.

[0011] Further, solving the multi-objective optimization model to obtain the optimal scheduling instruction set comprises the following steps: Based on the real-time operation state data and the preset scene classification rule, the current operation scene type is identified, based on the operation scene type, the historical optimal scheduling scheme and the weather forecast information, a plurality of candidate scheduling schemes are generated to generate a candidate scheduling scheme set, based on the operation scene type, the corresponding algorithm parameter combination is called from the preset parameter configuration library, the algorithm parameters include inertia weight, acceleration factor and constraint penalty coefficient; Based on the algorithm parameter combination, the candidate scheduling scheme is updated, the knowledge-guided mutation operation is performed on the updated candidate scheduling scheme, the fast non-dominated sorting method is used to sort the candidate scheduling scheme set after the mutation operation, the candidate scheduling scheme set is calculated by using the crowding distance calculation method, the step is repeatedly executed until the iteration termination condition is met, the iteration termination condition is that the maximum iteration number is reached or the improvement of the total operation cost of the candidate scheduling scheme is less than the preset threshold value for 10 consecutive iterations; The optimal scheduling scheme is selected from the finally obtained candidate scheduling scheme set, and the optimal scheduling instruction set is generated based on the optimal scheduling scheme.

[0012] Further, updating the candidate scheduling scheme based on the algorithm parameter combination comprises the following steps: For each candidate scheduling scheme, calculate the adjustment instruction, the adjustment instruction is composed of three parts of inertia, individual cognition and social learning, and the corresponding inertia weight, individual cognition weight and social learning weight are obtained from the preset parameter configuration library according to the current operation scene type; Add each candidate scheduling scheme to the adjustment instruction calculated in the previous step to obtain the updated candidate scheduling scheme, and perform constraint checking and amplitude limiting on the updated candidate scheduling scheme.

[0013] Further, the knowledge-guided mutation operation is performed on the updated candidate scheduling scheme, comprising the following steps: For each candidate scheduling scheme, detect whether the net charging and discharging amount of the battery in any continuous time period is greater than the preset threshold value, if so, use the power equal scaling method for deep mutation correction; Identify the start-up time point of the diesel engine in each candidate scheduling scheme, and extend the running time to the minimum start-up time or cancel the start-up instruction when the continuous running time is detected to be less than the preset time threshold; Identify whether the power generation of the renewable energy in each candidate scheduling scheme is greater than the time period of load demand, check the charging state of the battery in these time periods, charge the battery with the excess renewable energy when the battery is detected to be not charged, perform safety and feasibility verification on the candidate scheduling scheme after the mutation operation, roll back to the state before the mutation for the candidate scheduling scheme that fails the verification, and evaluate the total operation cost change of the candidate scheduling scheme before and after the mutation for the candidate scheduling scheme that passes the verification, and record the mutation type, mutation amplitude and mutation effect if the total operation cost becomes smaller, and establish a mutation knowledge base.

[0014] Further, the candidate scheduling scheme set calculated by using the crowding distance calculation method comprises the following steps: A non-dominated sorting algorithm is used to obtain a plurality of non-dominated layer sequences, each non-dominated layer has a plurality of candidate scheduling schemes, for the candidate scheduling schemes of each non-dominated layer, the candidate scheduling schemes are sorted according to each sub-cost, the crowding distance between the two candidate scheduling schemes located at the boundary is set to infinity, for the candidate scheduling schemes located at the non-boundary, the difference value of the sub-costs of the adjacent two candidate scheduling schemes is calculated, the difference value is normalized, the sum of the normalized difference values is calculated, the sum of the difference values is taken as the first distance of the candidate scheduling scheme, and the sum of the results obtained by adding all the first distances corresponding to the sub-costs is taken as the final crowding distance; The total number of the preset candidate scheduling scheme set is set, the candidate scheduling schemes are first selected according to the order from high to low of the non-dominated layers to join the candidate scheduling scheme set, and the candidate scheduling schemes are selected according to the order from large to small of the crowding distance in the same non-dominated layer, until the total number of the candidate scheduling scheme set reaches the preset number.

[0015] In a second aspect, the application provides an integrated power supply system, which comprises: A data acquisition module acquires micro-grid operation state data in real time, and the operation state data includes photovoltaic power generation power, wind power generation power, load demand power, diesel generator state, operation state, state of charge, temperature, voltage and current of the power supply equipment battery; A loss calculation module establishes a stress factor table of the battery, identifies and analyzes the effective charge and discharge cycles of the battery within a preset time window based on the current data of the battery, calculates the loss value within the time period of the effective charge and discharge cycles based on the stress factor table, converts the loss value into a battery life loss cost, and updates the stress factor table every preset time interval; The optimization scheduling module takes the total operation cost as an objective function, establishes an optimization model containing life loss cost, and the total operation cost includes diesel generator fuel cost, battery life loss cost, power purchase cost from the main power grid and renewable energy abandonment penalty cost, adopts a model predictive control framework, and obtains an optimal scheduling instruction set by solving the optimization model based on microgrid real-time operation state data and short-term prediction information in each optimization period by using an improved particle swarm optimization algorithm.

[0016] The beneficial effects of the present application are as follows: The present application accurately identifies and quantifies the life loss of the battery under random charging and discharging conditions by establishing a stress factor table and an improved Rainflow counting method, converts the abstract life decay into a calculable loss cost, includes the battery life loss cost in the optimization target, and realizes the balanced optimization of short-term operation and long-term equipment health together with the diesel fuel cost, power purchase cost, and energy abandonment penalty cost to form the total operation cost; the improved particle swarm optimization algorithm is used in combination with scene recognition, knowledge-guided mutation operation and non-dominated sorting to improve the feasibility and diversity of the scheduling scheme, adapt to the complex and changeable microgrid operation environment; the optimization of the battery charging and discharging strategy and the diesel engine start-stop logic reduces the abandonment of renewable energy, improves the utilization rate of clean energy, and reduces the dependence on traditional diesel power generation; the establishment of the loss prediction model and the mutation knowledge base realizes the dynamic updating of the stress factor table and the scheduling strategy, and improves the precision and robustness of the system in actual operation. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 An embodiment schematic diagram of the integrated power supply system scheduling method in the embodiments of the present application is shown in the figure. Figure 2 An embodiment schematic diagram of the integrated power supply system in the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0019] It can be understood that the terms "first", "second", etc. used in the present application can be used herein to describe various elements, but unless specifically stated, these elements are not limited by these terms. These terms are only used to distinguish the first element from another element. For example, without departing from the scope of the present application, the first xx script can be referred to as the second xx script, and similarly, the second xx script can be referred to as the first xx script.

[0020] As Figure 1As shown, a kind of integrated power system scheduling method, comprising: Step S1, real-time acquisition microgrid operating state data, operating state data include photovoltaic power generation, wind power generation, load demand, diesel generator state, power equipment battery operating state, state of charge, temperature, voltage and current.

[0021] Specifically, in the independent microgrid, the battery of power equipment is the key equipment to ensure system stability and improve new energy consumption, but its cost is high, and the life is limited, the existing microgrid energy management strategy, more focus on real-time power balance and operation cost optimization, but ignores the fine management of battery life, in wind light storage microgrid, due to the randomness of energy input and the volatility of load, the battery is in non-standard random charging and discharging condition for a long time. This condition will significantly accelerate the battery aging, in order to be able to evaluate and give priority to the life health of battery when carrying out energy scheduling of microgrid, realize the optimal life cycle cost of system, propose the present application.

[0022] First, in order to realize optimization decision, acquisition microgrid operating state data, operating state data include photovoltaic power generation, wind power generation, load demand power, diesel generator state, power equipment battery operating state, state of charge, temperature, voltage and current, it needs to be noted that, in the case that diesel generator state is opened, the power generation of diesel generator also needs to be collected, the operating state of power equipment battery refers to whether the battery is in charging state, discharging state or static state, the state of charge refers to the energy storage level of battery, the temperature, voltage and current data of battery can provide data basis for establishing battery prediction model.

[0023] Step S2, establish the stress factor table of battery, based on the current data of battery, the effective charging and discharging cycle of battery in preset time window is identified and analyzed, the loss value in the time period of effective charging and discharging cycle is calculated based on stress factor table, the loss value is converted into battery life loss cost, and the stress factor table is updated every preset time interval.

[0024] Specifically, in order to accurately calculate the life loss of the battery, a stress factor table of the battery is established in advance, the stress factor table stores the corresponding relationship between different charge and discharge cycle ranges and the capacity attenuation rate of the battery, can be established based on an accelerated aging experiment, based on the current data of the battery, the effective charge and discharge cycles of the battery within a preset time window are identified and analyzed, and the loss value within the time period of the effective charge and discharge cycles is calculated based on the stress factor table. The method of calculating the loss value will be explained in detail later. The loss value is converted into a battery life loss cost. Due to the different use of the battery, the stress factor table will also change with the situation, so the stress factor table is updated every preset time interval to improve the calculation accuracy of the loss value.

[0025] Step S3, taking the total operation cost as the objective function, an optimization model containing the life loss cost is established, the total operation cost includes the diesel generator fuel cost, the battery life loss cost, the purchase cost from the main power grid and the renewable energy abandonment penalty cost, using the model predictive control framework, based on the real-time operation state data and short-term prediction information of the microgrid in each optimization period, the improved particle swarm optimization algorithm is used to solve the optimization model to obtain the optimal scheduling instruction set.

[0026] Specifically, in order to consider the life loss of the battery when performing energy scheduling, taking the total operation cost as the objective function, an optimization model containing the life loss cost is established, wherein the total operation cost includes the diesel generator fuel cost, the battery life loss cost, the purchase cost from the main power grid and the renewable energy abandonment penalty cost. Using the model predictive control framework, based on the real-time operation state data and short-term prediction information of the microgrid in each optimization period, the state data includes photovoltaic power generation power, wind power generation power, battery remaining capacity, battery temperature, diesel engine generation state and load demand. The state data represents the current state of the system. The short-term prediction information includes photovoltaic power generation prediction, wind power generation prediction, load prediction and battery state prediction. The short-term prediction information can be obtained based on the corresponding prediction model established in advance. For example, the photovoltaic power generation prediction can be obtained through the pre-established photovoltaic power generation prediction model. Other short-term prediction models can also obtain corresponding prediction data through pre-established prediction models. The state data and short-term prediction information are input into the optimization model, and the improved particle swarm optimization algorithm is used to solve the multi-objective optimization model to obtain the optimal scheduling instruction set. The improved particle swarm optimization algorithm is an optimization algorithm fully considering the application scenario of actual microgrid scheduling. The specific solving process will be explained in detail later. The optimal scheduling instruction set includes diesel engine start-stop instructions, battery charge-discharge instructions, renewable energy limit generation instructions and load reduction instructions.

[0027] In this embodiment, calculating the loss value within the time period of the effective charge and discharge cycles includes the following steps: The extreme value analysis is performed on the current time series data of the battery, all local maximum points and local minimum points are identified, a peak-valley feature point sequence is formed, a complete cycle is identified based on the peak-valley feature point sequence, and the depth value of the complete cycle is calculated. The depth value of the identified complete cycle is matched and queried with the pre-established battery stress factor table to obtain the loss weight value corresponding to each depth range value. The loss weight values of each complete cycle are accumulated to obtain the loss value in the current analysis period.

[0028] Specifically, in order to extract key turning points from continuous current data, simplify data complexity, and lay a foundation for complete stress cycle identification, extreme value analysis is performed on the current time series data of the battery, all local maximum points and local minimum points are identified, the local maximum points are also called peak points (a point is higher than its adjacent points before and after it), and the local minimum points are also called valley points (a point is lower than its adjacent points before and after it). A complete cycle is identified based on the peak-valley feature point sequence, and the depth value of the complete cycle is calculated. The method for identifying complete cycles and calculating the depth value of the complete cycle will be explained in detail in the candidate. Then, the depth value of the identified complete cycle is matched and queried with the pre-established battery stress factor table, the physical cycle range value is converted into a standardized life loss weight, and the quantitative relationship between cycle stress and battery aging is established. That is, based on the electrochemical aging mechanism of the battery, a corresponding relationship table of different cycle depths and battery capacity decay rates is established through a large number of previous accelerated aging experiments. The actual identified cycle range is mapped to a standardized loss weight using a table lookup method or an interpolation method. Then, the loss weight values of each complete cycle are accumulated to obtain the loss value in the current analysis period.

[0029] In this embodiment, identifying a complete cycle based on the peak-valley feature point sequence includes the following steps: Four consecutive feature points are selected from the peak-valley feature point sequence at a time, the first difference value from the first feature point to the second feature point, the second difference value from the second feature point to the third feature point, and the third difference value from the third feature point to the fourth feature point are calculated. When the second difference value is less than or equal to the first difference value and less than or equal to the third difference value, it is judged that the second feature point to the third feature point constitutes a complete cycle, the second difference value is taken as the depth value of the complete cycle, and the feature points constituting the complete cycle are removed from the peak-valley feature point sequence. The peak-valley feature point sequence after removal is continuously identified for complete cycles until no new complete cycle can be identified.

[0030] Specifically, assuming that the obtained peak-valley feature point sequence is -50, -10, -40, +30, 10, 25, +5, for the continuous four peak-valley value points A: -50, B: -10, C: -40, D: +30, the difference between the adjacent two peak-valley value points is calculated to obtain 40, 30, and 70, and it is judged whether the fluctuation in the middle is less than the fluctuation on both sides. In the current example, 30 is less than 40 and also less than 70, which indicates that a complete charge-discharge cycle is formed from the second feature point to the third feature point, and the physical meaning is that the battery has experienced a discharge current of -10, which is relatively high, to a deeper discharge current of -40, and a return change in the range of 30A. This cycle causes a quantifiable life loss to the battery, and therefore it is indicated that the cycle has been completed. The second feature point and the third feature point are removed from the peak-valley feature point sequence, and then the peak-valley feature point sequence after removal is continuously identified for a complete cycle. At this time, the peak-valley feature point sequence is -50, +30, 10, 25, +5, and for the continuous four peak-valley value points -50, +30, 10, 25, the first difference from the first feature point to the second feature point, the second difference from the second feature point to the third feature point, and the third difference from the third feature point to the fourth feature point are calculated to obtain three differences, which are 80, 20, and 15, respectively. At this time, since 20 is less than 80 but greater than 15, it is indicated that the fluctuation from 30 to 10 is greater than the fluctuation from 10 to 25, which means that the current change is still continuing and has not formed a complete cycle of “peak-valley-peak” or “valley-peak-valley”. This fluctuation may be part of a larger cycle or more data points are needed to determine the cycle structure. At this time, these points are temporarily retained, and the corresponding physical meaning is that the current change mode of the battery is not clear enough to determine whether an independent stress cycle has been completed. Therefore, the subsequent data is continuously analyzed, that is, the next four data +30, 10, 25, +5 are analyzed. At this time, three differences are calculated to be 20, 15, and 20, respectively. At this time, the second difference is less than or equal to the first difference and less than or equal to the third difference, so it is indicated that a complete charge-discharge cycle is formed from the second feature point 10 to the third feature point 25, and the difference 15 between the two is taken as the depth value of the complete cycle. At this time, 10 and 25 are deleted from the peak-valley value feature point sequence to obtain -50, +30, +5. At this time, if there are subsequent data values, the remaining peak-valley value feature points can be continuously analyzed according to the method. If not, two complete cycles are identified from the current current data, and the depth values corresponding to the two complete cycles are 30 and 15. Then, the stress factor table is queried. Assuming that the stress factor table is 0-20 corresponding to a loss weight of 0.01, 20-50 corresponding to a loss weight of 0.05, 50-100 corresponding to a loss weight of 0.12, and 100 or more corresponding to a loss weight of 0.2, then the weight corresponding to 30 is 0.05 and the weight corresponding to 15 is 0.01. The corresponding calculated loss value is 0.5+0.1=0.06, which represents that the battery consumes 0.06 times of the standard cycle life under the working condition of this time period. This loss value can be used to calculate the life loss cost and be included in the optimization model. The above calculation of the random working condition life loss of the battery based on the improved Rainflow counting method can scientifically quantify the abstract life, and can ensure that the life loss evaluation of the battery under the random and intermittent working condition is more accurate, and can provide a more reliable basis for subsequent intelligent scheduling.

[0031] In this embodiment, updating the stress factor table includes the following steps: A battery loss prediction model is established, the input feature of the prediction model is the historical running state data of the battery, and the output feature is the loss prediction value of the battery in a future period of time; Based on the prediction model, the loss prediction value of the future period of time is obtained, the calculated loss value of the corresponding period of time is obtained, the deviation value of the two is obtained by comparing the calculated loss value and the loss prediction value, and if the deviation value exceeds the preset threshold, the stress factor table is adjusted based on the deviation value and the historical working condition of the battery.

[0032] Specifically, the above-mentioned loss value calculation method is an efficient algorithm based on real-time data, which can quickly identify the local peak and valley values in the battery charging and discharging current sequence, and accurately calculate the depth value and number of each stress cycle. This method relies on the current actual running data and can respond to the instantaneous changes of the battery in real time, such as sudden charging and discharging events, thereby providing timely life loss evaluation. However, the calculation of the loss value is directly related to the stress factor table. However, due to the complex and changeable actual use of the battery, the corresponding stress factor table of the battery may be different from the stress factor table obtained through previous experiments. In order to accurately calculate the loss of the battery, a battery loss prediction model is established based on historical battery operation data, including battery temperature, historical charging and discharging current, etc. The loss prediction value of the future preset period of time is predicted. Based on the prediction model, the loss prediction value of the future period of time is obtained, the calculated loss value of the corresponding period of time is obtained, the deviation value of the two is obtained by comparing the calculated loss value and the loss prediction value, and if the deviation value exceeds the preset threshold, it means that the calculated loss value is not accurate, which further indicates that the stress factor table has deviated. Therefore, the stress factor table is adjusted based on the deviation value. If the calculated loss value is less than the loss prediction value, and the system judges that the battery is in a long-term high-temperature, high-humidity, and frequent charging and discharging harsh working condition, the actual aging rate of the battery is much faster than the preset aging rate. Based on this result, the weight coefficient in the stress factor table is adjusted upward. If the calculated loss value is greater than the loss prediction value, the current battery working environment is relatively stable, the charging and discharging current is also long-term stable, and the aging is slow. Based on this result, the weight coefficient in the stress factor table is adjusted downward.

[0033] The above method makes the stress factor table used for calculating the loss value no longer mechanically apply fixed values, but actually perceive the actual health of the battery based on the prediction model, and adjust the stress factor table based on the prediction result, so that the calculated loss value is always of high accuracy.

[0034] In this embodiment, establishing a multi-objective optimization model including life loss cost includes the following steps: Define the decision variable, the decision variable includes diesel engine output power, battery charging and discharging power, renewable energy scheduling ratio, load reduction, set the target function as minimizing the total operation cost, set the constraint condition, the constraint condition includes power balance constraint, equipment operation constraint and battery life constraint, the battery life loss cost in the total cost is the battery life loss multiplied by the unit loss cost of the battery.

[0035] Specifically, the decision variable refers to the specific power instruction and state instruction that needs to be issued to each device in each decision cycle (such as every 5 minutes), wherein the diesel engine output power refers to how much power the diesel engine is instructed to output in the next scheduling cycle, the battery charging and discharging power refers to how much power the battery is instructed to charge or discharge in the next scheduling cycle, the renewable energy scheduling ratio refers to how much proportion of the power generated by photovoltaic power generation and wind power generation is used to supply the load, and how much proportion is abandoned, and the load reduction refers to how much non-critical load is instructed to cut off in the extreme case (such as when the energy is severely insufficient and the battery needs to be protected). The decision variable is an adjustable variable, which facilitates finding the best scheduling instruction for minimizing the total contract under the premise of meeting all constraint conditions. The constraint conditions include power balance constraint, equipment operation constraint and battery life constraint, wherein the power balance constraint refers to solar power + wind power + diesel power + battery discharge power = load power + battery charge power + abandoned power + load reduction power, the equipment operation constraint refers to the minimum and maximum output power limit of the diesel generator, the upper and lower limit of the battery energy storage, the battery charging and discharging power limit and the renewable energy scheduling limit.

[0036] In this embodiment, solving the multi-objective optimization model to obtain the optimal scheduling instruction set includes the following steps: Identify the current operation scene type based on real-time operation state data and pre-set scene classification rules, generate a candidate scheduling scheme set based on the operation scene type, historical high-quality scheduling scheme and weather forecast information, and call the corresponding algorithm parameter combination from the pre-set parameter configuration library based on the operation scene type, wherein the algorithm parameters include inertia weight, acceleration factor and constraint penalty coefficient. updating the candidate scheduling scheme based on the algorithm parameter combination, performing a knowledge-guided mutation operation on the updated candidate scheduling scheme, sorting the candidate scheduling scheme set after the mutation operation using a fast non-dominated sorting method, and calculating the candidate scheduling scheme set using a crowding distance calculation method, repeating the step until the iteration termination condition is met, the iteration termination condition being that the maximum number of iterations is reached or the improvement in the total running cost of the candidate scheduling scheme is less than a preset threshold for 10 consecutive iterations; selecting an optimal scheduling scheme from the final candidate scheduling scheme set, and generating an optimal scheduling instruction set based on the optimal scheduling scheme.

[0037] Specifically, in order to improve the convergence speed and improve the quality of the initial solution, different initialization strategies are adopted according to different micro-grid operation scenarios, real-time micro-grid operation state data is collected, the current operation scenario type is identified based on the operation state data and the preset scenario classification rule, the scenario classification rule is that the renewable energy output is greater than 120% of the load demand (wind and light sufficient scenario), the renewable energy output is less than 60% of the load demand (wind and light insufficient scenario), the load demand is greater than 85% of the rated capacity (peak scenario), and the key equipment fails (emergency standby scenario), and a plurality of candidate scheduling schemes are generated based on the operation scenario type, the historical high-quality scheduling scheme and the weather forecast information to generate a candidate scheduling scheme set, for example, in the low wind and light sufficient scenario, an initial scheme mainly charging the battery is generated; in the wind and light insufficient scenario, an initial scheme mainly powered by the diesel generator is generated, and the knowledge of the historical high-quality scheme is introduced to avoid generating obviously unreasonable initial solutions.

[0038] A parameter configuration library is established based on a large amount of experimental data, and an optimized parameter combination is obtained for each type of operation scenario threshold, wherein the inertia weight controls the balance between global exploration and local development ability, the acceleration factor affects the speed of particle movement to the optimal position, and the constraint penalty coefficient ensures the feasibility of the solution. The parameter values are verified by a large number of experiments to ensure the optimal performance in each scenario.

[0039] By executing the standard particle swarm optimization algorithm, the candidate scheduling scheme is updated based on the algorithm parameter combination, and the specific updating process will be explained in detail later. The knowledge-guided mutation operation on the updated candidate scheduling scheme means that based on the actual experience of micro-grid operation, a special mutation rule is designed. The battery charging and discharging depth mutation avoids overcharging and overdischarging, the diesel generator operation time mutation reduces frequent start and stop, and the renewable energy consumption mutation improves the utilization rate of clean energy. These mutation operations increase the guidance of domain knowledge based on the standard algorithm, and the specific mutation operation process will be explained in detail later.

[0040] In order to provide a rich and high-quality scheduling scheme, the candidate scheduling scheme set after the mutation operation is sorted using the fast non-dominated sorting method, the candidate scheduling scheme set is calculated using the crowding distance calculation method, the high-quality candidate scheduling scheme is effectively identified through the fast non-dominated sorting method, and the uniformity of the distribution of the candidate scheduling scheme in the candidate scheduling scheme set is ensured through the crowding distance.

[0041] The mutation operation of the candidate scheduling scheme and the operation of generating the candidate scheduling scheme set are repeatedly performed until the iteration termination condition is met, the iteration termination condition refers to reaching the maximum iteration number or the improvement of the total running cost of the candidate scheduling scheme in 10 consecutive iterations being less than a preset threshold, and finally, the optimal scheduling scheme is selected from the finally obtained candidate scheduling scheme set, and the optimal scheduling instruction set is generated based on the optimal scheduling scheme.

[0042] The above method is deeply optimized by specially aiming at the characteristics of the micro-grid scheduling problem, and provides effective technical support for the optimization scheduling decision of the intelligent integrated power supply equipment.

[0043] In the embodiment, updating the candidate scheduling scheme based on the algorithm parameter combination includes the following steps: For each candidate scheduling scheme, calculate the adjustment instruction, the adjustment instruction is weighted and synthesized by the inertia part, the individual cognitive part and the social learning part, and the corresponding inertia weight, individual cognitive weight and social learning weight are obtained from the preset parameter configuration library according to the current running scene type; Add each candidate scheduling scheme to the adjustment instruction calculated in the last step to obtain the updated candidate scheduling scheme, and perform constraint checking and amplitude limiting processing on the updated candidate scheduling scheme.

[0044] Specifically, in order to efficiently optimize each candidate scheduling scheme, for each candidate scheduling scheme, calculate the adjustment instruction, the adjustment instruction is weighted and synthesized by the inertia part, the individual cognitive part and the social learning part, wherein the inertia part refers to inheriting the adjustment instruction of the last candidate scheduling scheme, maintaining the continuity of the search, the weight is the inertia weight, the social learning part refers to the direction and distance of the adjustment to the current global optimal candidate scheduling scheme, the weight is the social learning weight, and the individual cognitive part refers to the direction and distance of the adjustment to the current candidate scheduling scheme itself historical optimal position, the weight is the individual cognitive weight. In order to generate a new generation of candidate scheduling scheme according to the adjustment instruction, add each candidate scheduling scheme to the adjustment instruction calculated in the last step to obtain the updated candidate scheduling scheme, update the candidate scheduling scheme, make the candidate scheduling scheme realize gradual improvement, and ensure the smoothness and physical realizability of the scheme update.

[0045] To ensure that the updated candidate scheduling scheme meets the basic operation requirements, constraint checking is performed on the updated candidate scheduling scheme, including: battery power constraint processing: ensuring that the charging and discharging power of the candidate scheduling scheme does not exceed the maximum allowed power; diesel generator output constraint processing: ensuring that the diesel generator output is between the minimum output and the rated power; power balance constraint processing: ensuring that the power generation of each candidate scheduling scheme is basically balanced with the load demand. The updated candidate scheduling scheme is also subjected to limiting processing to ensure that the adjustment range is within a reasonable range, specifically including: setting a maximum adjustment range, when the module of the adjustment instruction exceeds the maximum adjustment range, scaling the adjustment instruction in proportion, and the maximum adjustment range is dynamically adjusted according to the operation scene type, a larger value is set in the exploration stage, and a smaller value is set in the fine optimization stage.

[0046] The above method deeply combines the traditional particle swarm optimization algorithm with the specific characteristics of microgrid scheduling, solves the problems of insufficient interpretability and practicability of abstract algorithms in actual engineering applications, and provides reliable technical support for optimization decision of intelligent integrated power equipment.

[0047] In the embodiment, the knowledge-guided mutation operation performed on the updated candidate scheduling scheme includes the following steps: For each candidate scheduling scheme, it is detected whether the net charging and discharging amount of the battery in any continuous time period is greater than a preset threshold. If so, a power scaling method is used for deep mutation correction; The start time point of the diesel engine in each candidate scheduling scheme is identified, and when the continuous running time is detected to be less than a preset time threshold, the running time is extended to the minimum start time or the start instruction is cancelled; It is detected whether the power generation of the renewable energy in each candidate scheduling scheme is greater than the time period of the load demand, and the state of charge of the battery in these time periods is checked. When it is detected that the battery is not charged, the excess renewable energy is used to charge the battery; the safety and feasibility of the candidate scheduling scheme after the mutation operation is verified, the candidate scheduling scheme that fails the verification is rolled back to the state before the mutation, the total operation cost change of the candidate scheduling scheme before and after the mutation is evaluated for the candidate scheduling scheme that passes the verification, and if the total operation cost becomes smaller, the mutation type, mutation amplitude and mutation effect are recorded, and a mutation knowledge base is established.

[0048] Specifically, in order to prevent the overcharge and discharge of the battery, prolong the service life of the battery, based on the electrochemical characteristics of the battery, deep charge and discharge will accelerate the aging of the battery, through the sliding time window detection, the scheduling period that violates the depth limit of charge and discharge is identified. The equal proportion scaling method is used to maintain the time distribution characteristics of the original scheduling scheme, and only the safety limit is made in the amplitude, which not only ensures the safety of the candidate scheduling scheme, but also significantly prolongs the service life of the battery, prevents the premature failure of the battery caused by unreasonable scheduling scheme through preventive protection, and reduces the whole life cycle operation and maintenance cost.

[0049] In order to reduce the frequent start and stop of diesel generator, reduce equipment wear and operation cost, based on the mechanical characteristics of diesel generator, frequent start and stop will increase equipment wear and maintenance cost, through identifying short time operation instruction, using running time extension or starting cancellation two strategies for correction, extension strategy is suitable for important power supply period, cancellation strategy is suitable for alternative power supply period, ensure that the system reliability is not affected, through the above optimization operation strategy, not only improve the operation efficiency of diesel generator, but also reduce the equipment wear and maintenance cost.

[0050] In order to improve the utilization rate of renewable energy, reduce the abandonment, based on the principle of energy priority utilization, the storage charging is arranged in the period of renewable energy surplus, and the difference between wind and light output and load demand is detected in real time, the battery charging power is dynamically adjusted, and the minimum value is limited to ensure that the charging power does not exceed the equipment capacity and the total amount of available renewable energy, so as to realize the full and reasonable utilization of energy.

[0051] In order to ensure that the scheduling scheme after variation is safe and feasible, the safety and feasibility of the candidate scheduling scheme after variation operation is verified, the candidate scheduling scheme that fails to pass the verification is rolled back to the state before variation, and the total operation cost change of the candidate scheduling scheme before and after variation is evaluated for the candidate scheduling scheme that passes the verification, if the total operation cost is smaller, the variation type, variation amplitude and variation effect are recorded, the variation knowledge base is established, the experience is provided for subsequent variation operation through the establishment of variation knowledge base, and the self learning optimization of variation strategy is realized.

[0052] In the embodiment, the candidate scheduling scheme set obtained by using the crowding distance calculation method includes the following steps: The non dominated sorting algorithm is used to obtain a plurality of non dominated layer sequences, each non dominated layer has a plurality of candidate scheduling schemes, for the candidate scheduling scheme of each non dominated layer, the candidate scheduling schemes are sorted according to each sub cost, the crowding distance between the two candidate scheduling schemes located at the boundary is set to infinity, for the candidate scheduling scheme located at the non boundary, the difference value of the sub cost of the adjacent two candidate scheduling schemes is calculated, the difference value is normalized, the sum of the normalized difference value is calculated, the difference value is taken as the first distance of the candidate scheduling scheme, and the sum of the first distances corresponding to all sub costs is taken as the final crowding distance. The total number of the preset candidate scheduling scheme set is first selected according to the order from high to low of the non-dominated layer, and the candidate scheduling scheme is added into the candidate scheduling scheme set, and in the same non-dominated layer, the candidate scheduling scheme is selected according to the order from large to small of the crowding distance, until the total number of the candidate scheduling scheme set reaches the preset value.

[0053] Specifically, assuming that there are 7 candidate scheduling schemes in a non-dominated layer, taking only fuel cost and battery life loss cost as an example, scheme A: (80, 60), scheme B: (70, 75), scheme C: (90, 50), scheme D: (85, 65), scheme E: (75, 80) and scheme F: (95, 45), the order of fuel cost is: scheme B: 70, scheme E: 75, scheme A: 80, scheme D: 85, scheme C: 90, scheme F: 95, and the order of battery life loss cost is: scheme F: 45, scheme C: 50, scheme A: 60, scheme D: 65, scheme B: 75, scheme E: 80, the range of cost 1 is 95-70=25, the range of cost 2 is 80-45=35, for the sorting sequence [B, E, A, D, C, F] of cost 1, scheme B and scheme F are boundary schemes, and the crowding distance is set to infinity, the first distance of scheme E is: (the difference between B and E) / 25+(the difference between E and A) / 25=10 / 24=0.4, and the first distances of schemes A, D and C are calculated in the same way, which are 0.4 respectively, and the first distances of each scheduling scheme are calculated according to the battery life loss cost, which are scheme F: infinity, scheme C: 0.429, scheme A: 0.429, scheme D: 0.429, scheme B: 0.429, scheme E: infinity, and the result of adding the first distances corresponding to all sub-costs is used as the final crowding distance of each scheme, and the final crowding distance result is scheme A: 0.829, scheme B: infinity, scheme C: 0.829, scheme D: 0.829, scheme E: infinity and scheme F: infinity.

[0054] The total number of the preset candidate scheduling scheme set is first selected according to the order from high to low of the non-dominated layer, and the candidate scheduling scheme is added into the candidate scheduling scheme set, and in the same non-dominated layer, the candidate scheduling scheme is selected according to the order from large to small of the crowding distance, until the total number of the candidate scheduling scheme set reaches the preset value.

[0055] Further, the multiple non-dominated layer sequences obtained by using the non-dominated sorting algorithm include the following steps: For any two candidate scheduling schemes A and B, the rule for judging the dominance relationship is: if all the objective function values of scheme A are not worse than scheme B, and at least one objective function value is strictly better than scheme B, then scheme A dominates scheme B; record the set of other schemes dominated by each candidate scheduling scheme, and the number of schemes dominated by other schemes, select the schemes with zero dominated number from all candidate scheduling schemes to form the first layer of non-dominated solution set, reduce the dominated number of the schemes dominated by the selected non-dominated solution set by one, and select the schemes with zero dominated number from the remaining schemes to form the next layer of non-dominated solution set, repeat the above process until all candidate scheduling schemes are assigned to the corresponding non-dominated layer.

[0056] Specifically, it is assumed that there are 5 candidate scheduling schemes, and two objectives need to be minimized: fuel cost (F1) and life loss cost (F2). The scheme data is as follows: scheme A: F1=100, F2=50, scheme B: F1=80, F2=70, scheme C: F1=120, F2=40, Scheme D: Fl=90, F2=80, Scheme E: Fl=110, F2=60, calculate dominated quantity: Scheme A: dominated by B (B's Fl=80<100, F2=70>50, not satisfy domination condition), dominated by C (C's F2=40<50, but Fl=120>100, not satisfy), dominated by D (D's Fl=90<100, F2=80>50, not satisfy), dominated by E (E's Fl=110>100, F2=60>50, not satisfy), dominated quantity=0; Scheme B: compared with A (Fl=80<100, F2=70>50, not dominate A), compared with C (Fl=80<120, F2=70>40, not dominate C), compared with D (Fl=80<90, F2=70<80, dominate D), compared with E (Fl=80<110, F2=70>60, not dominate E), dominated quantity=0; Scheme C: compared with A (Fl=120>100, F2=40<50, not dominate A), compared with B (Fl=120>80, F2=40<70, not dominate B), compared with D (Fl=120>90, F2=40<80, not dominate D), compared with E (Fl=120>110, F2=40<60, dominate E), dominated quantity=0; Scheme D: dominated by B, compared with A (Fl=90<100, F2=80>50, not dominate), compared with C (Fl=90<120, F2=80>40, not dominate), compared with E (Fl=90<110, F2=80>60, not dominate), dominated quantity=1; Scheme E: dominated by C, compared with A (Fl=110>100, F2=60>50, not dominate), compared with B (Fl=110>80, F2=60<70, not dominate), compared with D (Fl=110>90, F2=60<80, not dominate), dominated quantity=1, after the above steps, get the first non-dominated layer: Scheme A, B, C (dominated quantity=0), the second non-dominated layer: Scheme D, E (dominated quantity=1).

[0057] The above describes a dispatching method of an integrated power supply system in an embodiment of the present application, and the following describes an integrated power supply system in an embodiment of the present application, please refer to Figure 2 An embodiment of an integrated power supply system in an embodiment of the present application includes: A data acquisition module, which acquires micro-grid operation state data in real time, the operation state data including photovoltaic power generation power, wind power generation power, load demand power, diesel generator state, operation state, state of charge, temperature, voltage and current of the power supply equipment storage battery; The loss calculation module establishes a stress factor table of the battery, identifies and analyzes the effective charge-discharge cycles of the battery within a preset time window based on current data of the battery, calculates a loss value within a time period of the effective charge-discharge cycles based on the stress factor table, converts the loss value into a battery life loss cost, and updates the stress factor table every preset time interval; The optimization scheduling module takes the total operation cost as an objective function, establishes an optimization model containing the life loss cost, the total operation cost including a diesel generator fuel cost, a battery life loss cost, a power purchase cost from a main power grid and a renewable energy abandonment penalty cost, adopts a model predictive control framework, and obtains an optimal scheduling instruction set by solving the optimization model based on real-time operation state data and short-term prediction information of the microgrid in each optimization period.

[0058] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present disclosure.

[0059] The above embodiments only express several implementation manners of the present application, and the description is specific and detailed, but it should not be understood as a limitation on the scope of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims. The above is only a preferred embodiment of the present application, and does not limit the present application. Any modification, equivalent replacement and improvement within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. An integrated power system scheduling method, characterized in that, include: Step S1: Collect microgrid operation status data in real time. The operation status data includes photovoltaic power generation, wind power generation, load demand power, diesel generator status, power equipment battery operation status, state of charge, temperature, voltage and current. Step S2: Establish a stress factor table for the battery. Based on the battery's current data, identify and analyze the effective charge-discharge cycles of the battery within a preset time window. Calculate the loss value within the effective charge-discharge cycle time period based on the stress factor table, convert the loss value into battery life loss cost, and update the stress factor table every preset time interval. Step S3: With minimizing the total operating cost as the objective function, establish an optimization model that includes lifetime loss costs. The total operating cost includes diesel generator fuel costs, battery lifetime loss costs, electricity purchase costs from the main grid, and renewable energy abandonment penalty costs. Using a model predictive control framework, within each optimization cycle, based on the microgrid's real-time operating status data and short-term forecast information, use an improved particle swarm optimization algorithm to solve the optimization model and obtain the optimal scheduling instruction set.

2. The method according to claim 1, characterized in that, Calculating the loss value over the effective charge-discharge cycle period includes the following steps: Extreme value analysis is performed on the current time series data of the battery to identify all local maxima and local minima, forming a peak-valley feature point sequence. Based on the peak-valley feature point sequence, complete cycles are identified, and the depth value of the complete cycle is calculated. The depth value of the identified complete cycle is matched and queried with a pre-established battery stress factor table to obtain the loss weight value corresponding to each depth range. The loss weight values ​​of each complete cycle are accumulated to obtain the loss value in the current analysis time period.

3. The method according to claim 2, characterized in that, The complete loop identification based on peak-valley feature point sequence includes the following steps: Four consecutive feature points are selected at once from the peak-valley feature point sequence. The first difference between the first feature point and the second feature point, the second difference between the second feature point and the third feature point, and the third difference between the third feature point and the fourth feature point are calculated respectively. When the second difference is less than or equal to the first difference and less than or equal to the third difference, it is determined that the second feature point and the third feature point form a complete cycle. The second difference is used as the depth value of the complete cycle. The feature points that form the complete cycle are removed from the peak-valley feature point sequence. The complete cycle is identified again for the removed peak-valley feature point sequence until no new complete cycle can be identified.

4. The method according to claim 1, characterized in that, Updating the stress factor table includes the following steps: Establish a battery loss prediction model. The model takes the historical operating status data of the battery as input and outputs the predicted loss value of the battery over a future period of time. The predicted loss value for a future time period is obtained based on the prediction model. The calculated loss value for the corresponding time period is obtained. The calculated loss value and the predicted loss value are compared to obtain the deviation value between the two. If the deviation value exceeds the preset threshold, the stress factor table is adjusted based on the deviation value and the historical operating conditions of the battery.

5. The method according to claim 1, characterized in that, Establishing a multi-objective optimization model that includes lifetime loss costs includes the following steps: Define decision variables, including diesel engine output power, battery charging and discharging power, renewable energy dispatch ratio, and load reduction. Set the objective function to minimize total operating cost. Set constraints, including power balance constraints, equipment operation constraints, and battery life constraints. The battery life loss cost in the total cost is the battery life loss amount multiplied by the battery unit loss cost.

6. The method according to claim 1, characterized in that, Solving the multi-objective optimization model to obtain the optimal scheduling instruction set includes the following steps: Based on real-time operation status data and preset scenario classification rules, the current operation scenario type is identified. Based on the operation scenario type, historical high-quality scheduling schemes and weather forecast information, multiple candidate scheduling schemes are generated to form a candidate scheduling scheme set. Based on the operation scenario type, the corresponding algorithm parameter combination is called from the preset parameter configuration library. The algorithm parameters include inertia weight, acceleration factor and constraint penalty coefficient. The candidate scheduling scheme is updated based on the combination of algorithm parameters. Knowledge-guided mutation operation is performed on the updated candidate scheduling scheme. The candidate scheduling scheme set after mutation is sorted using the fast non-dominated sorting method. The candidate scheduling scheme set is calculated using the congestion distance calculation method. This step is repeated until the iteration termination condition is met. The iteration termination condition is that the maximum number of iterations is reached or the improvement of the total running cost of the candidate scheduling scheme in 10 consecutive iterations is less than a preset threshold. The optimal scheduling scheme is selected from the final set of candidate scheduling schemes, and the optimal scheduling instruction set is generated based on the optimal scheduling scheme.

7. The method according to claim 6, characterized in that, Updating candidate scheduling schemes based on algorithm parameter combinations includes the following steps: For each candidate scheduling scheme, an adjustment instruction is calculated. The adjustment instruction is a weighted synthesis of three parts: inertia, individual cognition, and social learning. The corresponding inertia weight, individual cognition weight, and social learning weight are obtained from the preset parameter configuration library according to the current running scenario type. Each candidate scheduling scheme is added to the adjustment instruction calculated in the previous step to obtain the updated candidate scheduling scheme. Constraint checks and amplitude limiting are then performed on the updated candidate scheduling scheme.

8. The method according to claim 6, characterized in that, Performing a knowledge-guided mutation operation on the updated candidate scheduling scheme includes the following steps: For each candidate scheduling scheme, check whether the net charge and discharge amount of the battery within any consecutive time period is greater than the preset threshold. If so, use the power scaling method to perform deep variation correction. Identify the start time of the diesel engine in each candidate scheduling scheme. When the continuous running time is detected to be less than the preset time threshold, extend the running time to the minimum start time or cancel the start command. Identify the time periods in each candidate scheduling scheme where the power generation of renewable energy exceeds the load demand, check the charging status of the battery during these time periods, and if the battery is not charged, use the excess renewable energy to charge the battery. The candidate scheduling schemes after mutation are verified for safety and feasibility. For candidate scheduling schemes that fail the verification, they are rolled back to their state before mutation. For candidate scheduling schemes that succeed in the verification, the change in the total operating cost before and after the mutation is evaluated. If the total operating cost decreases, the mutation type, mutation magnitude and mutation effect are recorded, and a mutation knowledge base is established.

9. The method according to claim 6, characterized in that, The process of calculating the candidate scheduling scheme set using the congestion distance method includes the following steps: A non-dominated sorting algorithm is used to obtain multiple non-dominated layer sequences. Each non-dominated layer has multiple candidate scheduling schemes. For each candidate scheduling scheme in a non-dominated layer, the candidate scheduling schemes are sorted according to each sub-cost. The congestion distance between two candidate scheduling schemes located at the boundary is set to infinity. For candidate scheduling schemes located at non-boundaries, the difference between its sub-cost and that of its two adjacent candidate scheduling schemes is calculated. After normalizing the difference, the sum of the normalized difference is calculated. The sum of the difference is used as the first distance of the candidate scheduling scheme. The result of adding the first distances corresponding to all sub-costs is used as the final congestion distance. The total number of candidate scheduling schemes is preset. First, candidate scheduling schemes are selected and added to the candidate scheduling scheme set in descending order of non-dominated layer. Within the same non-dominated layer, candidate scheduling schemes are selected in descending order of congestion distance until the total number of candidate scheduling schemes reaches the preset value.

10. An integrated power supply system for executing an integrated power supply system scheduling method, characterized in that, The system includes the following modules: The data acquisition module collects real-time microgrid operating status data, including photovoltaic power generation, wind power generation, load demand power, diesel generator status, power equipment battery operating status, state of charge, temperature, voltage and current. The loss calculation module establishes a stress factor table for the battery, identifies and analyzes the effective charge-discharge cycles of the battery within a preset time window based on the battery's current data, calculates the loss value within the effective charge-discharge cycle time period based on the stress factor table, converts the loss value into battery life loss cost, and updates the stress factor table every preset time interval. The optimization scheduling module takes minimizing the total operating cost as the objective function and establishes an optimization model that includes lifetime loss costs. The total operating cost includes diesel generator fuel costs, battery lifetime loss costs, electricity purchase costs from the main grid, and renewable energy abandonment penalty costs. A model predictive control framework is adopted. In each optimization cycle, based on the real-time operating status data of the microgrid and short-term forecast information, an improved particle swarm optimization algorithm is used to solve the optimization model to obtain the optimal scheduling instruction set.

Images