Micro-grid multi-objective optimization scheduling method considering service life of energy storage battery

Abstract

The invention relates to a micro-grid multi-objective optimization scheduling method considering the service life of an energy storage battery, and belongs to the field of active power distribution network operation analysis. The invention relates to a micro-grid multi-objective optimization scheduling method considering the service life of an energy storage battery. The method comprises the following steps: establishing an energy storage system refined operation model in a wind-light storage micro-grid; on the premise of maximum income and maximum new energy consumption, the method has the beneficial effects that the new energy power output fluctuation is responded in time, so that the charge and discharge control of the energy storage system is efficient and accurate, and the purposes of improving the energy storage utilization efficiency, the charge and discharge control accuracy and the power output stability are achieved; according to the improved multi-objective grey wolf optimization algorithm, an initial population strategy, a population individual iteration updating strategy and a multi-objective optimal solution solving mechanism are improved, and the convergence speed, the solution set diversity and the calculation efficiency of the multi-objective grey wolf optimization algorithm are improved through multi-strategy cooperation; the problems of excessive uncertainty and single target of the existing wind and light storage micro-grid are effectively solved.

Description

Technical Field

[0001] This invention relates to a multi-objective optimization scheduling method for microgrids that takes into account the lifespan of energy storage batteries, and belongs to the field of active distribution network operation analysis. Background Technology

[0002] Against the backdrop of a escalating global energy crisis and environmental problems, renewable energy has developed rapidly, playing a significant role in addressing climate change and ensuring power supply. However, its randomness and volatility lead to adverse effects from large-scale grid integration, with prominent issues such as wind and solar power curtailment. Energy storage, with its rapid power regulation and storage capabilities, has become a key means to mitigate power fluctuations and reduce waste in renewable energy sources. Wind-solar-storage microgrids have attracted attention as an efficient and clean energy supply method, but achieving low-carbon, stable operation and coordinated control using energy storage is a key research focus. Energy storage batteries offer flexible regulation and can mitigate the intermittent and volatile effects of renewable energy, but they also suffer from rapid lifespan depletion and high costs. Currently, establishing energy storage lifespan models and optimizing scheduling to extend lifespan and maximize economic value is a research hotspot. However, existing methods suffer from drawbacks in cost estimation, such as underestimating lifespan depletion costs, low estimation accuracy, overestimating depletion costs, or difficulty in balancing computational accuracy and solution speed. It is necessary to establish reasonable models that balance the economics and solution speed of energy storage scheduling and assess cycle life to balance lifespan loss and capacity utilization.

[0003] Traditional single-time-scale scheduling struggles to efficiently coordinate the diverse adjustable resources within a microgrid. Furthermore, day-ahead scheduling suffers from long sampling intervals and low prediction accuracy, making it difficult to meet real-time power balance requirements. Model predictive control (MPC), a mature control strategy, is applied to microgrid optimization scheduling, reducing operating costs and addressing voltage fluctuations. However, its coordination with energy storage in wind-solar-storage microgrids requires further exploration. In wind-solar-storage microgrid systems, multi-objective scheduling strategy formulation is crucial, involving economic operation, system stability, and environmental performance. Strategy formulation must comprehensively consider the initial cost of the energy storage system, its lifetime degradation, and the cost of interaction with the external grid. Swarm intelligence optimization algorithms have become a research hotspot for solving such problems. For example, this invention, CN111340299B, discloses a multi-objective optimization scheduling method for microgrids, including determining the microgrid structure and uncertainty modeling; establishing load models and default models based on different load characteristics; considering microgrid cost optimization and user satisfaction maximization while performing multi-objective modeling; and using an improved NSGA-II algorithm for solution. The beneficial effects of this invention are as follows: First, by improving the generation of crossover operator coefficients in the NSGA-II algorithm, the complexity of the algorithm is effectively reduced, and the convergence speed and accuracy are further improved. Second, the scheduling method that comprehensively considers the overall operating cost of the microgrid and user satisfaction significantly improves user satisfaction compared to considering the economics of the microgrid alone, thus meeting actual electricity demand. However, there are technical problems such as difficulty in responding to fluctuations in the output of new energy power in a timely manner, resulting in inefficient and inaccurate charging and discharging control of the energy storage system and low energy storage utilization efficiency. Summary of the Invention

[0004] To address the problem of inefficient and inaccurate charge and discharge control and low energy storage utilization efficiency caused by the difficulty in responding to fluctuations in new energy power output, this invention proposes a multi-objective optimization scheduling method for microgrids that takes into account the lifespan of energy storage batteries.

[0005] The technical solution adopted by the present invention to solve the above problems is as follows: A multi-objective optimization scheduling method for microgrids considering energy storage battery lifetime is proposed. This method establishes a refined operation model of the energy storage system in a wind-solar-storage microgrid. Under the premise of maximizing revenue and renewable energy consumption, it assesses energy storage lifetime loss through charge / discharge depth and cycle count. Utilizing the refined energy storage operation model and MPC theory, a multi-timescale scheduling architecture for the microgrid considering energy storage lifetime loss is established. Through asymptotic multi-timescale timing, renewable energy output information is effectively tracked, and intraday rolling corrections are further performed based on day-ahead decisions. An improved version of the traditional Grey Wolf (GWO) algorithm is proposed to realize a multi-objective Grey Wolf optimization algorithm, achieving a smooth transition from global exploration to local exploitation and enhancing the algorithm's ability to escape local optima. Finally, Pareto optimal solutions for various objective functions are searched and measured through archiving and leader selection mechanisms. Furthermore, the refined operation model of the energy storage system in the wind-solar-storage microgrid includes a correlation characteristic model between the cycle life of the energy storage system and the depth of charge and discharge. Under the premise of maximizing the benefits of the energy storage charging and discharging strategy and maximizing the absorption of new energy, the energy storage life loss is evaluated by the depth of charge and discharge and the number of cycles.

[0006] Furthermore, the relationship between the cycle life of the battery energy storage and the depth of discharge in the energy storage system is obtained by power function fitting, i.e. (1) In the formula: The number of cycles required for the battery to reach the retirement condition; The number of charge-discharge cycles when storing energy in a battery at 100% depth of discharge; Depth of discharge for battery energy storage charge-discharge cycles; As the fitting factor, and All use the factory-fixed parameters of the battery energy storage equipment.

[0007] Furthermore, regarding the correlation characteristic model of the charge / discharge depth, the daily lifespan cost of the battery depends on its usage that day. Therefore, the battery cycle count and discharge depth are first quantified to estimate the maintenance cost of battery lifespan. Under ideal operating conditions, the battery can complete one full charge / discharge cycle during scheduling, and its calculated... The coefficient is: (2) In the formula: This represents the equivalent cycle number under ideal operating conditions, assuming the battery completes one full charge and discharge cycle per day. =100%), then the equivalent number of cycles per day is ; It is also necessary to convert the number of cycles consumed by the battery at different discharge depths at different times into the equivalent number of cycles at 100% discharge depth, and the corresponding conversion factor. for: (3) In the formula: This represents the equivalent number of cycles under ideal operating conditions; for Cycle lifetime at constant depth of discharge; This is a time period measured in hours; This refers to the number of hours per day; The battery cycle conversion factor obtained under standard working days The battery's working life can then be calculated. and the degree of battery life loss The battery's service life can be calculated by dividing the number of cycles at 100% depth of discharge by the number of cycles converted to a typical workday, and then dividing by the number of days in a year. (4) (5) After adjusting the capital recovery factor, the average daily depreciation cost is: (6) In the formula: The annual interest rate; Annual battery maintenance costs, Daily average loss cost Piecewise linearization is performed to accelerate the solution speed of the optimal scheduling. The expression for the daily battery depreciation cost after piecewise linearization is: (7) In the formula: for The change in battery depth of discharge over a period of time; and The parameters are obtained by piecewise linearization at different discharge depths based on equation (5).

[0008] Furthermore, the determination method for energy storage charging and discharging strategies involves the following steps: S1: If the wind and solar power generation of the microgrid is higher than the local load, the excess wind and solar power will be used to charge the energy storage device first. If there is still a surplus, it will be sold back to the grid to generate revenue. S2: If the wind and solar power generation of the microgrid is lower than the local load, the energy storage device adopts different charging and discharging strategies according to the time-of-use electricity price. s21: When electricity prices are at their peak, priority should be given to selling electricity to generate revenue. Energy storage devices should discharge as much as possible within permissible limits to meet electricity demand. When energy storage devices cannot meet demand, it is necessary to consider purchasing electricity from the grid and triggering the demand response mechanism to reduce electricity costs as much as possible. s22: When the electricity price is in a low-price period, the power purchase strategy should be given priority. Within the charge status range of the energy storage device, try to purchase electricity from the large power grid. At this time, the purchased power can simultaneously meet the charging needs of the energy storage device and the local load gap. s23: When electricity prices are at a stable level, consider from multiple perspectives whether it is necessary to purchase electricity.

[0009] Furthermore, in each iteration of the traditional Grey Wolf Algorithm (GWO), the best, second-best, and third-best solutions in the population are respectively set as leaders. , and The lowest-ranking gray wolf The location will be based on , and Update according to the following formula:

[0010] In the formula: This represents the current iteration number; For the individual gray wolf in the first The new position after the next iteration; , and They are respectively from , and The resulting displacement; , and They are respectively , and Location; , and For algorithm parameters, and A random number located in the interval [0,1]. The maximum number of iterations, The improved multi-objective gray wolf optimization algorithm enhances its optimization and convergence performance by refining the initialization population strategy, thereby strengthening its local exploitation and global exploration capabilities. Specifically, it assumes a wolf pack size of... Search dimensions are The position of the gray wolf in the wolf pack can be represented as follows: The Tent chaotic map can leverage its sensitivity and randomness to combine with optimization algorithms to quickly find a better solution in the search space, as shown below: (15) After generating the sequence, it is mapped to the solution space: (16) In the formula: , To define the upper and lower bounds of the variable, we generate a population of 1 / 2 size. Define the optimal point parameter and select the smallest prime number. Best Points Collection Coordinates of the points: (17) In the formula: {•} indicates taking the decimal part. Mapping to the solution space: (18) In the formula: and For the first The upper and lower bounds of the dimension are used to generate the remaining half population size.

[0011] Furthermore, the multi-objective gray wolf optimization algorithm guides the algorithm from coarse-grained search to fine-grained search by dynamically adjusting the leader wolf weight, introducing random perturbations from Levy's flight, and modifying the convergence factor. The calculation formula is as follows: To weaken the dominance of the leader wolf in the later stages of iteration, a dynamic weight coefficient that decreases linearly with the number of iterations is designed: (19) In the formula: The maximum number of iterations is preset. , , They represent , , The wolves' weights are initially assigned to 0.5, 0.3, and 0.2 based on their importance. The weight coefficients are... , , Follow Linear decrease makes the algorithm effective in the early stages ( The leader wolf guides the search direction. ), and later ( The weights approach 0, reducing the herd mentality. Introducing the Lévy flight stochastic term, its step size is calculated using the following formula: (20) Where: scaling factor (0.01); Lévy distribution index ; Independently follows a standard normal distribution The random variable; Γ(•) is the Gamma function used to calculate the scale parameter. σ , Combining dynamic weights and Lévy flight, the individual position update formula is defined as: (twenty one) By using the linear convergence factor in the traditional gray wolf optimization algorithm The strategy was changed to a nonlinear convergence factor. Its expression is: (twenty two) Multi-objective algorithm design: Set the mesh expansion coefficient Grid size coefficient The archive has a maximum number of members. When the archive is full, a roulette wheel algorithm is activated as shown below to delete old solutions from grids with congested search space: (twenty three) (twenty four) In the formula: For the first The selection probability of each grid cell For the first The cumulative probability of each grid cell; An array to record the number of solutions in the grid; The number of grid cells containing solutions; It is a constant parameter.

[0012] Furthermore, the microgrid multi-timescale scheduling architecture considering energy storage lifetime loss includes the MPC multi-timescale optimization scheduling architecture and the wind-solar-storage microgrid multi-timescale optimization scheduling model. The MPC multi-timescale optimization includes day-ahead and intraday states. In the day-ahead stage, based on the day-ahead long-scale distributed renewable energy forecast results and day-ahead load forecast information, economic and environmental scheduling optimization is performed with a time resolution of 1 hour and a scheduling cycle of 24 hours. During the optimization process, the trade-off between the benefits and losses caused by different charging and discharging depths of the energy storage system is considered. The scheduling plan is then distributed from the scheduling center to each unit, and the day-ahead demand response is used as a fixed value in the intraday rolling scheduling stage. Rolling optimization and feedback correction are performed with a time resolution of 15 minutes, a scheduling cycle of 4 hours, and a time interval of 15 minutes. Each rolling operation only executes the first period plan, and the current actual output value of the system is used as the initial value for the new round of rolling optimization scheduling, forming a closed-loop control. The scheduling result deviation caused by the forecast error is corrected in a timely manner through the continuous updating of ultra-short-term forecast information.

[0013] Furthermore, in the multi-timescale optimal scheduling model for the wind-solar-storage microgrid, the day-ahead stage is characterized by economical and environmentally friendly scheduling, with the objective function as follows: (25) (26) (27) (28) In the formula: the superscript DA indicates that the variable is a variable for the intraday period; the previous day period h, with a time resolution of 1h; The cost of purchasing electricity from the main power grid; Cost of responding to demand; This refers to the daily depreciation cost of the battery. Daily operating cost of microgrid; This refers to the amount of wind and solar power that has been curtailed. for The wind power curtailment caused by scheduling restrictions during a certain period is the difference between the theoretical maximum wind power and the actual power used. for Wasted power during specific time periods; The unit time-of-use electricity price for purchasing electricity from the main power grid; The active power purchased by the microgrid from the main grid; This represents the compensation cost for interruptible loads. This represents the amount of active power interruption calls for interruptible loads. The objective function for the intraday phase is: (29) (30) Control variables in the real-time phase: (31) The update strategy expression for each control variable during the real-time rolling optimization phase: (32) In the formula: the superscript ID indicates that the variable is a variable for the intraday phase; This is the current rolling scheduling time; For finite time-domain scheduling intervals; To quickly adjust the deviation between the resource output value and the current day's planned value; This refers to the deviation control variables in the current dispatch cycle during the intraday phase, including the active and reactive power of the large power grid. / Active / reactive power of energy storage system charging and discharging Distributed photovoltaic and wind power reactive power ; To adjust costs in real time; for Power that needs to be temporarily purchased or sold from the grid during a certain period due to power imbalance; This is a penalty coefficient for the rate of change of energy storage power, used to suppress frequent charge-discharge switching; for The difference between the energy storage capacity of a given period and that of the previous period; To control the initial running state of the variables; Decision variables that are control variables; This indicates taking the L2 norm of a vector; These are the weighting coefficients corresponding to each adjustment resource deviation item.

[0014] Furthermore, it is necessary to systematically constrain the multi-timescale optimal scheduling model of the wind-solar-storage microgrid. This can be achieved through joint constraints based on power balance constraints, distributed renewable energy output constraints, and electrochemical energy storage device constraints, ensuring the stability of the multi-timescale optimal scheduling model. These constraints are specifically expressed by the following formulas: (33) In the formula: for Real-time system load power, (34) In the formula: For new energy Efforts made at all times; For new energy installed capacity, (35) In the formula: , These are the inverter's rated charging power and rated discharging power, respectively. State of charge of the energy storage device; , These represent the upper and lower limits of the state of charge of the energy storage device.

[0015] The beneficial effects of this invention are: 1. Timely response to fluctuations in new energy power output enables efficient and precise control of energy storage system charging and discharging, and high energy storage utilization efficiency, thereby achieving the goals of improving energy storage utilization efficiency, charging and discharging control precision, and power output stability; 2. An improved multi-objective gray wolf optimization algorithm is proposed. Compared with the traditional algorithm, the initial population strategy, the population individual iterative update strategy, and the multi-objective optimal solution-finding mechanism are improved. Through multi-strategy collaboration, the convergence speed, solution set diversity, and computational efficiency of the multi-objective gray wolf optimization algorithm are improved. 3. It effectively solves the problems of excessive uncertainty and single objective in the current wind, solar, and energy storage microgrids. Attached Figure Description

[0016] Figure 1 This is a diagram of the energy storage charging and discharging strategy of the present invention; Figure 2 This is a flowchart of the improved MOGWO of the present invention; Figure 3 This is a schematic diagram of the wind-solar-storage microgrid system structure of the present invention; Figure 4 This is a schematic diagram of the multi-timescale optimization scheduling framework of the present invention. Detailed Implementation

[0017] Specific Implementation Method 1: A Multi-Objective Optimization Scheduling Method for Microgrids Considering Energy Storage Battery Lifespan: This method focuses on the scheduling optimization of wind-solar-storage microgrids. It constructs a refined energy storage operation model, builds a multi-timescale scheduling architecture based on Model Predictive Control (MPC) theory, and improves the traditional Grey Wolf Algorithm (GWO) to achieve multi-objective optimization. Ultimately, it achieves efficient and stable operation of the microgrid. By establishing a refined operation model of the energy storage system in the wind-solar-storage microgrid, the characteristics of energy storage can be accurately characterized. It deeply considers various physical and operational characteristics of the energy storage system, accurately simulates the charging and discharging behavior of energy storage devices under different operating conditions, and provides a solid foundation for the formulation of subsequent scheduling strategies. This ensures that scheduling decisions are more aligned with the actual operating conditions of the energy storage system, providing a basis for supporting multi-objective optimization. It provides a basic model framework for scheduling under the premise of maximizing revenue and renewable energy consumption, enabling accurate evaluation of the impact of different scheduling strategies on revenue and renewable energy consumption based on this model, thus providing strong support for multi-objective optimization. Under the premise of maximizing both revenue and renewable energy consumption, this paper assesses energy storage lifetime loss by evaluating charge / discharge depth and cycle count. Utilizing a refined energy storage operation model and MPC theory, a multi-timescale dispatch architecture for microgrids that considers energy storage lifetime loss is established. This approach balances economic and environmental goals, pursuing both maximum revenue and maximum renewable energy consumption. While ensuring the economic benefits of the microgrid, it actively promotes the utilization of renewable energy, reduces dependence on traditional energy sources, and aligns with the requirements of sustainable development. Assessing energy storage lifetime loss through charge / discharge depth and cycle count allows for a more scientific consideration of the lifespan of energy storage systems, preventing premature damage due to overuse and reducing the operating and maintenance costs of the microgrid. Furthermore, the multi-timescale dispatch architecture established based on the refined energy storage operation model and MPC theory fully utilizes information from different time scales. Day-ahead decisions provide macro-level dispatch plans for the microgrid, while intraday rolling corrections dynamically adjust based on real-time information, enabling the dispatch strategy to better adapt to the uncertainty of renewable energy output and improving dispatch accuracy and adaptability. By progressively tracking renewable energy output information across multiple time scales, and further performing intraday rolling corrections based on day-ahead decisions, more accurate and detailed renewable energy output information can be gradually obtained. This enables effective tracking of renewable energy output, which helps microgrids better utilize renewable energy, improve the renewable energy absorption rate, and reduce wind and solar curtailment. Intraday rolling corrections based on day-ahead decisions can promptly address fluctuations and uncertainties in renewable energy output. When actual output deviates from day-ahead forecasts, intraday rolling corrections can quickly adjust dispatch strategies to ensure the stable operation of the microgrid and reduce risks caused by output uncertainty.

[0018] Because multi-objective optimization problems are prone to getting trapped in local optima, resulting in solutions that are not truly Pareto optimal, an improved multi-objective gray wolf optimization algorithm (MOGWO) is proposed to address the traditional Gray Wolf Algorithm (GWO). This improved algorithm achieves a smooth transition from global exploration to local exploitation, enhancing its ability to escape local optima. The improved MOGWO achieves a smooth transition from global exploration to local exploitation, better balancing the capabilities of global and local searches during the search process. Global exploration helps discover a wider range of potential solution spaces and find possible regions of global optima; local exploitation can delve deeper into these regions, improving solution accuracy and thus more effectively finding the optimal solution. Enhancing the algorithm's ability to escape local optima is the key to the improvement. The improved algorithm better avoids this situation, increasing the probability of finding a global Pareto optimal solution and providing a better solution for microgrid scheduling. By employing archiving and leader selection mechanisms, the algorithm searches for Pareto optimal solutions across multiple objective functions. The archiving mechanism preserves non-dominated solutions found during the search process, preventing their loss in subsequent searches and ensuring the algorithm comprehensively searches for Pareto optimal solutions for various objective functions. This provides a rich set of options for microgrid dispatching, allowing decision-makers to select the most suitable dispatching strategy from these Pareto optimal solutions based on actual needs. The leader selection mechanism selects a suitable leader from the archive according to certain rules, guiding the search process in a more promising direction. This mechanism improves the algorithm's search efficiency, reduces unnecessary searches, and enables the algorithm to find high-quality Pareto optimal solutions more quickly, meeting the real-time and accuracy requirements of microgrid dispatching. In summary, through a series of model building, algorithm improvement and scheduling architecture design, the proposed solution has significant advantages in improving the revenue of wind-solar-storage microgrids, promoting the consumption of new energy sources, extending the lifespan of energy storage, and optimizing scheduling strategies, and can effectively improve the operating efficiency and stability of microgrids.

[0019] Specific Implementation Method Two: A refined operation model for energy storage systems in wind-solar-storage microgrids includes a model of the correlation between the cycle life and depth of charge / discharge of the energy storage system. Since the lifespan of electrochemical energy storage is closely related to its operating mode, influencing factors include temperature, discharge rate, depth of discharge, and the number of overcharges and over-discharges. In practical applications, heat dissipation devices are used to control the temperature of the energy storage equipment, thereby limiting the discharge rate within a reasonable range. Under a low cycle frequency of daily single charge and discharge, electrochemical energy storage can generally last until its expected lifespan. However, the superposition of frequent charge / discharge cycles and low state of charge will severely shorten the battery's expected lifespan. Therefore, short-term planning is considered in scheduling, i.e., the impact of the number of charging cycles and the depth of charge / discharge in a day on the battery's total lifespan. The relationship between the cycle life and depth of discharge of the battery energy storage is obtained by fitting a power function, i.e. (1) In the formula: The number of cycles required for the battery to reach the retirement condition; The number of charge-discharge cycles when storing energy in a battery at 100% depth of discharge; Depth of discharge for battery energy storage charge-discharge cycles; As the fitting factor, and All use the factory-fixed parameters of the battery energy storage equipment.

[0020] Specific Implementation Method 3: Equation (1) shows that as the DOD increases, the battery life cycle decreases non-linearly. Therefore, the battery cycle count and discharge depth are quantified using a charge / discharge depth correlation characteristic model. The daily lifespan cost of the battery depends on its usage that day. First, the battery cycle count and discharge depth are quantified to estimate the maintenance cost of the battery life. Under ideal conditions, the battery can complete one full charge / discharge cycle during scheduling, and its calculated... The coefficient is: (2) In the formula: This represents the equivalent cycle number under ideal operating conditions, assuming the battery completes one full charge and discharge cycle per day. =100%), then the equivalent number of cycles per day is ; In actual operation, to maximize economic benefits or new energy consumption, the optimal scheduling result of energy storage equipment may not be a one-way full charge and discharge cycle carried out on weekdays. The cost loss caused by irregular charging and discharging of energy storage equipment cannot be simply derived from equation (2). It is also necessary to convert the number of cycles consumed by the battery at different discharge depths into the equivalent number of cycles at 100% discharge depth, and the corresponding conversion factor. for: (3) In the formula: This represents the equivalent number of cycles under ideal operating conditions; for Cycle lifetime at constant depth of discharge; This is a time period measured in hours; This refers to the number of hours per day; The battery cycle conversion factor obtained under standard working days The battery's working life can then be calculated. and the degree of battery life loss The battery's service life can be obtained by dividing the number of cycles at 100% discharge depth by the number of cycles converted to a typical working day, and then dividing by the number of days in a year. (4) (5) (4) (5) Since batteries can operate normally for 10 years or more, and taking inflation into account, the initial investment cost of the battery is directly factored in. Divide by years of operation This is unreasonable and therefore needs to be corrected using a capital recovery factor. The adjusted capital recovery factor results in the following daily depreciation cost: (6) In the formula: The annual interest rate; Annual battery maintenance costs, Equations (2) to (6) reveal that, except for DOD as a decision variable, all other quantities are either given or derived from DOD, regarding the average daily depreciation cost. Piecewise linearization is performed to accelerate the solution speed of the optimal scheduling. The expression for the daily battery depreciation cost after piecewise linearization is as follows: (7) In the formula: for The change in battery depth of discharge over a period of time; and The parameters are obtained by piecewise linearization at different discharge depths based on equation (5).

[0021] Specific implementation method four: such as Figure 1 As shown, the determination method for the energy storage charging and discharging strategy described in this embodiment includes the following steps: S1: If the wind and solar power generation of the microgrid is higher than the local load, the excess wind and solar power will be used to charge the energy storage device first. If there is still a surplus, it will be sold back to the grid to generate revenue. S2: If the wind and solar power generation of the microgrid is lower than the local load, the energy storage device adopts different charging and discharging strategies according to the time-of-use electricity price. s21: When electricity prices are at their peak, priority should be given to selling electricity to generate revenue. Energy storage devices should discharge as much as possible within permissible limits to meet electricity demand. When energy storage devices cannot meet demand, it is necessary to consider purchasing electricity from the grid and triggering the demand response mechanism to reduce electricity costs as much as possible. s22: When the electricity price is in a low-price period, the power purchase strategy should be given priority. Within the charge status range of the energy storage device, try to purchase electricity from the large power grid. At this time, the purchased power can simultaneously meet the charging needs of the energy storage device and the local load gap. S23: When electricity prices are stable, consider from multiple perspectives whether it is necessary to purchase electricity; By taking into account the charging and discharging of energy storage based on the load of the storage device and the wind and solar power generation of the microgrid, as well as the current electricity price, we can effectively determine whether to purchase electricity from external sources and maximize the cost savings of the overall energy storage charging and discharging strategy.

[0022] Specific Implementation Method 5: In each iteration of the traditional Gray Wolf Algorithm (GWO), the best, second-best, and third-best solutions in the population are respectively set as leaders. , and The lowest-ranking gray wolf The location will be based on , and Update according to the following formula:

[0023] In the formula: This represents the current iteration number; For the individual gray wolf in the first The new position after the next iteration; , and They are respectively from , and The resulting displacement; , and They are respectively , and Location; , and For algorithm parameters, and A random number located in the interval [0,1]. The maximum number of iterations, Because traditional gray wolf optimization algorithms randomly generate individuals within a given range when generating the initial population, the randomness and uncertainty of the initial individuals are relatively large. Therefore, this paper improves the optimization and convergence performance of the algorithm by improving the initial population strategy, and enhances the local development and global exploration capabilities. Wind-solar-storage microgrids are complex, multi-variable, high-dimensional systems. For high-dimensional complex systems, this paper improves the ergodicity by generating some initial individuals using a Tent chaotic sequence, and obtains a uniform set of points with minimal deviation whose order of deviation is not affected by the dimension of the solution space through a set of optimal points. This provides superior theoretical support for solving high-dimensional problems. The improved multi-objective gray wolf optimization algorithm is as follows: assuming the wolf pack size is... Search dimensions are The position of the gray wolf in the wolf pack can be represented as The Tent chaotic map can leverage its sensitivity and randomness to combine with optimization algorithms to quickly find a better solution in the search space, as shown below: (15) After generating the sequence, it is mapped to the solution space: (16) In the formula: , To define the upper and lower bounds of the variable, we generate a population of 1 / 2 size. Define the optimal point parameter and select the smallest prime number. Best Points Collection Coordinates of the points: (17) In the formula: {•} indicates taking the decimal part. Mapping to the solution space: (18) In the formula: and For the first The upper and lower bounds of the dimension are used to generate the remaining half population size.

[0024] Specific implementation method six: such as Figure 2 As shown, the Gray Wolf Optimization Algorithm (GWO) described in this embodiment is prone to getting trapped in local optima when solving complex multimodal optimization problems, mainly due to the decrease in population diversity and the leader wolf (GWO) in the later stages of the algorithm. , , To address the over-guidance of the leader wolf algorithm, a balanced approach between global exploration and local exploitation is achieved by integrating dynamic weights with the individual update strategy of Lévy flight. Dynamically adjusting the leader wolf weight guides the algorithm from coarse-grained search to fine-grained search, preventing premature convergence. Introducing random perturbations from Lévy flight enhances the algorithm's ability to escape local optima. Furthermore, the convergence factor is modified. The calculation formula better coordinates the algorithm's exploration and convergence capabilities. To mitigate the dominance of the leader wolf in later iterations, a dynamic weight coefficient that decreases linearly with the number of iterations is designed. (19) In the formula: The maximum number of iterations is preset. , , They represent , , The wolves' weights are initially assigned to 0.5, 0.3, and 0.2 based on their importance. The weight coefficients are... , , Follow Linear decrease makes the algorithm effective in the early stages ( The leader wolf guides the search direction. ), and later ( The weights approach 0, reducing the herd mentality. Introducing the Lévy flight stochastic term, its step size is calculated using the following formula: (20) In the formula: the scaling factor (0.01) is used to suppress the absolute magnitude of the Lévy step size and prevent random perturbations from damaging the convergence stability; The Lévy distribution index controls the long-tail property of the step size distribution, generating a step size distribution with significant long-tail characteristics, balancing local search and global jump. Independently follows a standard normal distribution The random variable; Γ(•) is the Gamma function used to calculate the scale parameter. σ , Combining dynamic weights and Lévy flight, the individual position update formula is defined as: (twenty one) By optimizing the linear convergence factor in the traditional Grey Wolf (GWO) algorithm The strategy was changed to a nonlinear convergence factor. Its expression is: (twenty two) By setting the mesh expansion coefficient in the multi-objective gray wolf optimization algorithm Grid size coefficient , Multi-objective algorithm design: Set the mesh expansion coefficient Grid size coefficient The archive has a maximum number of members. When the archive is full, a roulette wheel algorithm is activated as shown below to delete old solutions from grids with congested search space: (twenty three) (twenty four) In the formula: For the first The selection probability of each grid cell For the first The cumulative probability of each grid cell; An array to record the number of solutions in the grid; The number of grid cells containing solutions; It is a constant parameter.

[0025] Specific implementation method seven: such as Figure 3 and Figure 4 As shown, the wind-solar-storage microgrid system structure described in this embodiment includes an integrated wind-solar-storage microgrid system, which is an advanced power system that integrates wind power, photovoltaic power generation, and energy storage technologies. Its efficient and stable operation relies on precise system control, especially when implementing multi-objective dispatch strategies. The control system plays a crucial role in this process. The core of the multi-objective dispatch strategy lies in balancing multiple conflicting objectives, such as economy, power supply reliability, and environmental protection. In the integrated wind-solar-storage microgrid, these objectives are typically achieved through the charging and discharging strategies of the energy storage system and its interaction with the external power grid. The control system needs to monitor the power generation of the photovoltaic panels and wind turbines in real time. Based on weather conditions and to predict future power generation, the control system can adjust the charging and discharging strategies of the energy storage system to ensure that solar and wind energy can be utilized to the maximum extent during peak wind and solar power generation periods. At the same time, when renewable energy is insufficient, the system can rely on the energy storage system and the main power grid to provide continuous power supply. Distributed wind and solar power generation uses the control module to provide feedback on energy flow to sensitive loads, adjustable loads, and general loads. Then, it is transmitted to the distribution network through the upper-level dispatch management system. The upper-level dispatch management system can also control the distribution network to transmit energy flow to the energy storage system for storage, as well as to sensitive loads, adjustable loads, and general loads. The aforementioned microgrid multi-timescale scheduling architecture considering energy storage lifetime loss includes the MPC multi-timescale optimized scheduling architecture and the wind-solar-storage microgrid multi-timescale optimized scheduling model. Since the core of MPC lies in utilizing the latest acquired predicted state information to implement continuous rolling optimized control, power prediction is the foundation for achieving optimized scheduling. However, prediction itself is not the focus of this invention. Therefore, this paper uses source-load prediction data as known information input to the model. Therefore, this paper uses source-load forecast data as known information input into the model. The multi-timescale optimization of MPC includes day-ahead and intraday states. The day-ahead stage is the economically optimal and renewable energy consumption optimal scheduling. By considering time-of-use pricing, the operation decisions and lifetime losses of energy storage system charging and discharging, the system achieves economic optimization. Through energy storage regulation, the system achieves renewable energy consumption optimization. The intraday stage is the deviation correction stage. Through rolling optimization and the establishment of a closed-loop feedback mechanism, the power balance and minimum real-time adjustment cost under real-time fluctuations of renewable energy power are achieved. In the day-ahead stage, based on the day-ahead long-scale distributed renewable energy forecast results and day-ahead load forecast information, with a time resolution of 1 hour and 24 hours... To optimize economic and environmental dispatching during the dispatch cycle, the trade-off between the benefits and losses arising from different charging and discharging depths of the energy storage system is considered during the optimization process. This dispatch plan is then distributed from the dispatch center to each unit. Because the optimization process considers the trade-off between the benefits and losses arising from different charging and discharging depths of the energy storage system, the objective function is to minimize the total operating cost. Considering the integration of renewable energy, the objective function is to maximize the proportion of renewable energy integration. The day-ahead demand response is used as a fixed value in the intraday rolling dispatching stage. Furthermore, because the forecast timescale in the day-ahead stage is relatively long, the increased uncertainties affecting renewable energy and load forecasting lead to lower forecast accuracy. The accuracy is low, and the large deviation between renewable energy output and load forecasts makes it difficult to meet the requirements of power balance. Therefore, the day-ahead demand response is used as a fixed value in the intraday rolling dispatch stage. Rolling optimization and feedback correction are carried out with a time resolution of 15 minutes, a dispatch cycle of 4 hours, and a time interval of 15 minutes. Each rolling operation only executes the first period plan, and the current actual output value of the system is used as the initial value for the new round of rolling optimization dispatch, forming a closed-loop control. By continuously updating the ultra-short-term forecast information, the dispatch result deviation caused by the forecast error is corrected in a timely manner, which effectively improves the predictability of the impact of the long forecast time scale of the day-ahead stage on renewable energy and load forecasts.

[0026] In each rolling optimization window, ultra-short-term forecasts are made on renewable energy output and load information for the future window period. The objective function is to minimize the expected output deviation of each regulating resource and the real-time adjustment cost. Each rolling operation executes only the first period's plan, while using the system's current actual output value as the initial value for the new round of rolling optimization scheduling, forming a closed-loop control. By continuously updating the ultra-short-term forecast information, the scheduling result deviation caused by forecast errors is corrected in a timely manner, thereby improving the microgrid's optimized scheduling accuracy while ensuring economic efficiency and environmental friendliness.

[0027] Specific Implementation Method Eight: In the multi-time-scale optimal scheduling model for the wind-solar-storage microgrid, the day-ahead stage is economical and environmentally friendly scheduling, and the objective function is as follows: (25) (26) (27) (28) In the formula: the superscript DA indicates that the variable is a variable for the intraday period; the previous day period h, with a time resolution of 1h; The cost of purchasing electricity from the main power grid; Cost of responding to demand; This refers to the daily depreciation cost of the battery. Daily operating cost of microgrid; This refers to the amount of wind and solar power that has been curtailed. for The wind power curtailment caused by scheduling restrictions during a certain period is the difference between the theoretical maximum wind power and the actual power used. for Wasted power during specific time periods; The unit time-of-use electricity price for purchasing electricity from the main power grid; The active power purchased by the microgrid from the main grid; This represents the compensation cost for interruptible loads. This represents the amount of active power interruption calls for interruptible loads. Through day-ahead scheduling optimization, the system's economic efficiency and environmental friendliness can be maximized. Because the real-time power generation of renewable energy within the system is constantly changing, the optimization results obtained in the daytime phase cannot meet the accuracy requirements of the real-time system. Therefore, when MPC is introduced into the real-time phase, the scheduling instructions are narrowed down to a shorter time scale to satisfy the deterministic requirements of the optimization results. The objective function for the intraday phase is: The objective function for the intraday phase is: (29) (30) Control variables in the real-time phase: (31) The update strategy expression for each control variable during the real-time rolling optimization phase: (32) In the formula: the superscript ID indicates that the variable is a variable for the intraday phase; This is the current rolling scheduling time; For finite time-domain scheduling intervals; To quickly adjust the deviation between the resource output value and the current day's planned value; This refers to the deviation control variables in the current dispatch cycle during the intraday phase, including the active and reactive power of the large power grid. / Active / reactive power of energy storage system charging and discharging Distributed photovoltaic and wind power reactive power ; To adjust costs in real time; for Power that needs to be temporarily purchased or sold from the grid during a certain period due to power imbalance; This is a penalty coefficient for the rate of change of energy storage power, used to suppress frequent charge-discharge switching; for The difference between the energy storage capacity of a given period and that of the previous period; To control the initial running state of the variables; Decision variables that are control variables; This indicates taking the L2 norm of a vector; These are the weighting coefficients corresponding to each adjustment resource deviation item; The aforementioned intraday phase is the deviation adjustment phase. Based on the forecast information updated by ultra-short-term forecasts, the optimization objective is to minimize the deviation between the rapid adjustment resource output value and the day-ahead planned value and to minimize the real-time adjustment cost. Combining the day-ahead part of the algorithm, it can be determined that the adjustable resources in the intraday adjustment phase have a smaller weighting coefficient, which means that the adjustment priority is higher and the system is more inclined to use the resource to smooth out the power fluctuations of renewable energy during intraday adjustment.

[0028] Specific Implementation Method Nine: A systematic constraint is also needed on the multi-timescale optimization scheduling model of the wind-solar-storage microgrid. This constraint is achieved through joint constraints of power balance, distributed renewable energy output, and electrochemical energy storage devices, ensuring the stability of the multi-timescale optimization scheduling model. These constraints are specifically expressed by the following formulas: (33) In the formula: for Real-time system load power, (34) In the formula: For new energy Efforts made at all times; For new energy installed capacity, (35) In the formula: , These are the inverter's rated charging power and rated discharging power, respectively. State of charge of the energy storage device; , These are the upper and lower limits of the state of charge of the energy storage device, respectively. Through the calculations of equations (33) to (35), excellent stability performance can be provided for the operation of the multi-time-scale optimization scheduling model of wind-solar-storage microgrid, ensuring that the electrical power, electrochemical energy storage equipment and distributed new energy output are all within the normal operating range during the operation of the model, thus ensuring both equipment stability and operational saturation.

[0029] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent substitutions, and improvements made to the above embodiments without departing from the scope of the present invention, based on the technical essence of the present invention and within the spirit and principles of the present invention, shall still fall within the protection scope of the present invention.

Claims

1. A multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries, characterized in that: This paper establishes a refined operation model for energy storage systems in wind-solar-storage microgrids. Under the premise of maximizing revenue and renewable energy consumption, it assesses energy storage lifetime loss through charge / discharge depth and cycle count. Using the refined energy storage operation model and MPC theory, a multi-timescale scheduling architecture for microgrids considering energy storage lifetime loss is established. Through asymptotic multi-timescale timing, renewable energy output information is effectively tracked, and intraday rolling corrections are further performed based on day-ahead decisions. An improved multi-objective gray wolf optimization algorithm is proposed to realize a smooth transition from global exploration to local development, enhancing the algorithm's ability to escape local optima. Pareto optimal solutions for various objective functions are searched and measured through archiving and leader selection mechanisms.

2. The microgrid multi-objective optimization scheduling method considering energy storage battery life as described in claim 1, characterized in that: The refined operation model of the energy storage system in the wind-solar-storage microgrid includes a correlation characteristic model of the energy storage system's cycle life and depth of charge / discharge. The relationship between the cycle life of the battery energy storage and the depth of discharge in the energy storage system's cycle life is obtained by power function fitting, i.e. (1) In the formula: The number of cycles required for the battery to reach the retirement condition; The number of charge-discharge cycles when storing energy in a battery at 100% depth of discharge; Depth of discharge for battery energy storage charge-discharge cycles; As the fitting factor, and All use the factory-fixed parameters of the battery energy storage equipment.

3. The microgrid multi-objective optimization scheduling method considering energy storage battery life as described in claim 2, characterized in that: The correlation model of charge / discharge depth shows that the daily lifespan cost of a battery depends on its usage that day. Therefore, the battery cycle count and discharge depth are first quantified to estimate the maintenance cost of battery lifespan. Under ideal conditions, a battery can complete one full charge / discharge cycle during scheduling, and its calculated... The coefficient is: (2) In the formula: This represents the equivalent cycle number under ideal operating conditions, assuming the battery completes one full charge and discharge cycle per day. =100%), then the equivalent number of cycles per day is ; It is also necessary to convert the number of cycles consumed by the battery at different discharge depths at different times into the equivalent number of cycles at 100% discharge depth, and the corresponding conversion factor. for: (3) In the formula: This represents the equivalent number of cycles under ideal operating conditions; for Cycle life at constant depth of discharge; This is a time period measured in hours; This refers to the number of hours per day; The battery cycle conversion factor obtained under standard working days The battery's working life can then be calculated. and the degree of battery life loss The battery's service life can be obtained by dividing the number of cycles at 100% discharge depth by the number of cycles converted to a typical working day, and then dividing by the number of days in a year. (4) (5) After adjusting the capital recovery factor, the average daily depreciation cost is: (6) In the formula: The annual interest rate; Annual battery maintenance costs, Daily average loss cost Piecewise linearization is performed to accelerate the solution speed of the optimal scheduling. The expression for the daily battery depreciation cost after piecewise linearization is as follows: (7) In the formula: for Changes in battery depth of discharge over time; and The parameters are obtained by piecewise linearization at different discharge depths based on equation (5).

4. The multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries according to claim 3, characterized in that: The determination of energy storage charging and discharging strategies is carried out through the following steps: S1: If the wind and solar power generation of the microgrid is higher than the local load, the excess wind and solar power will be used to charge the energy storage device first. If there is still a surplus, it will be sold back to the grid to generate revenue. S2: If the wind and solar power generation of the microgrid is lower than the local load, the energy storage device adopts different charging and discharging strategies according to the time-of-use electricity price. s21: When electricity prices are at their peak, priority should be given to selling electricity to generate revenue. Energy storage devices should discharge as much as possible within permissible limits to meet electricity demand. When energy storage devices cannot meet demand, it is necessary to consider purchasing electricity from the grid and triggering the demand response mechanism to reduce electricity costs as much as possible. s22: When the electricity price is in a low-price period, the power purchase strategy should be given priority. Within the charge status range of the energy storage device, try to purchase electricity from the large power grid. At this time, the purchased power can simultaneously meet the charging needs of the energy storage device and the local load gap. s23: When electricity prices are at a stable level, consider from multiple perspectives whether it is necessary to purchase electricity.

5. A multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries according to claim 1, characterized in that: In the traditional Grey Wolf Algorithm (GWO) described above, the best, second-best, and third-best solutions in the population are respectively set as leaders in each iteration. , and The lowest gray wolf The location will be based on , and Update according to the following formula: In the formula: This represents the current iteration number; For the individual gray wolf in the first The new position after the next iteration; , and They are respectively from , and The resulting displacement; , and They are respectively , and Location; , and For algorithm parameters, and A random number located in the interval [0,1]. The maximum number of iterations, The improved multi-objective gray wolf optimization algorithm enhances its optimization and convergence performance by refining the initialization population strategy, thereby strengthening its local exploitation and global exploration capabilities. Specifically, it assumes a wolf pack size of... Search dimensions are The position of the gray wolf in the wolf pack can be represented as The Tent chaotic map can leverage its sensitivity and randomness to combine with optimization algorithms to quickly find a better solution in the search space, as shown below: (15) After generating the sequence, it is mapped to the solution space: (16) In the formula: , To define the upper and lower bounds of the variable, we generate a population of 1 / 2 size. Define the optimal point parameter and select the smallest prime number. Best Points Collection Coordinates of the points: (17) In the formula: {•} indicates taking the decimal part. Mapping to the solution space: (18) In the formula: and For the first The upper and lower bounds of the dimension are used to generate the remaining half population size.

6. A multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries, as described in claim 5, is characterized in that: The multi-objective gray wolf optimization algorithm described above guides the algorithm from coarse-grained search to fine-grained search by dynamically adjusting the leader wolf weight, introducing random perturbations from Levy's flight, and modifying the convergence factor. The calculation formula is as follows: To weaken the dominance of the leader wolf in the later stages of iteration, a dynamic weight coefficient that decreases linearly with the number of iterations is designed: (19) In the formula: The maximum number of iterations is preset. , , They represent , , The wolves' weights are initially assigned to 0.5, 0.3, and 0.2 based on their importance. The weight coefficients are... , , Follow Linear decrease makes the algorithm effective in the early stages ( The leader wolf guides the search direction. ), and later ( The weights approach 0, reducing the herd mentality. Introducing the Lévy flight stochastic term, its step size is calculated using the following formula: (20) Where: scaling factor (0.01); Lévy distribution index ; Independently follows a standard normal distribution The random variable; Γ(•) is the Gamma function used to calculate the scale parameter. σ , Combining dynamic weights and Lévy flight, the individual position update formula is defined as: (21) By using the linear convergence factor in the traditional gray wolf optimization algorithm The strategy was changed to a nonlinear convergence factor. Its expression is: (22) By setting the mesh expansion coefficient in the multi-objective gray wolf optimization algorithm Grid size coefficient , Multi-objective algorithm design: Set the mesh expansion coefficient Grid size coefficient The archive has a maximum number of members. When the archive is full, a roulette wheel algorithm is activated as shown below to delete old solutions from grids with congested search space: (23) (24) In the formula: For the first The selection probability of each grid cell For the first The cumulative probability of each grid cell; An array to record the number of solutions in the grid; The number of grid cells containing solutions; It is a constant parameter.

7. A multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries according to claim 1, characterized in that: The aforementioned microgrid multi-timescale scheduling architecture considering energy storage lifetime loss includes the MPC multi-timescale optimization scheduling architecture and the wind-solar-storage microgrid multi-timescale optimization scheduling model. The MPC multi-timescale optimization includes day-ahead and intraday states. In the day-ahead stage, based on the day-ahead long-scale distributed renewable energy forecast results and day-ahead load forecast information, economic and environmental scheduling optimization is performed with a time resolution of 1 hour and a scheduling cycle of 24 hours. During the optimization process, the trade-off between the benefits and losses caused by different charging and discharging depths of the energy storage system is considered. The scheduling plan is then distributed from the scheduling center to each unit, and the day-ahead demand response is used as a fixed value to enter the intraday rolling scheduling stage. Rolling optimization and feedback correction are performed with a time resolution of 15 minutes, a scheduling cycle of 4 hours, and a time interval of 15 minutes. Each rolling operation only executes the first period plan, and the current actual output value of the system is used as the initial value for the new round of rolling optimization scheduling, forming a closed-loop control. The scheduling result deviation caused by the forecast error is corrected in a timely manner through the continuous updating of ultra-short-term forecast information.

8. A multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries, as described in claim 7, characterized in that: In the multi-time-scale optimal scheduling model for the wind-solar-storage microgrid, the day-ahead stage is characterized by economic and environmentally friendly scheduling, with the objective function as follows: (25) (26) (27) (28) In the formula: the superscript DA indicates that the variable is a variable for the intraday period; the previous day period h, with a time resolution of 1h; The cost of purchasing electricity from the main power grid; Cost of responding to demand; This refers to the daily depreciation cost of the battery. Daily operating cost of microgrid; This refers to the amount of wind and solar power that has been curtailed. for The wind power curtailment caused by scheduling restrictions during a certain period is the difference between the theoretical maximum wind power and the actual power used. for Wasted power during specific time periods; The unit time-of-use electricity price for purchasing electricity from the main power grid; The active power purchased by the microgrid from the main grid; This represents the compensation cost for interruptible loads. This represents the amount of active power interruption calls for interruptible loads. The objective function for the intraday phase is: (29) (30) Control variables in the real-time phase: (31) The update strategy expression for each control variable during the real-time rolling optimization phase: (32) In the formula: the superscript ID indicates that the variable is a variable for the intraday phase; This is the current rolling scheduling time; For finite time-domain scheduling intervals; To quickly adjust the deviation between the resource output value and the current day's planned value; This refers to the deviation control variables in the current dispatch cycle during the intraday phase, including the active and reactive power of the large power grid. / Active / reactive power of energy storage system charging and discharging Distributed photovoltaic and wind power reactive power ; To adjust costs in real time; for Power that needs to be temporarily purchased or sold from the grid during a certain period due to power imbalance; This is a penalty coefficient for the rate of change of energy storage power, used to suppress frequent charge-discharge switching; for The difference between the energy storage power of the current period and the previous period; To control the initial running state of the variables; Decision variables that are control variables; This indicates taking the L2 norm of a vector; These are the weighting coefficients corresponding to each adjustment resource deviation item.

9. A multi-objective optimization scheduling method for microgrids considering the lifespan of energy storage batteries, as described in claim 8, characterized in that: Furthermore, systematic constraints need to be imposed on the multi-timescale optimization scheduling model of the wind-solar-storage microgrid. These constraints are jointly imposed through power balance constraints, distributed renewable energy output constraints, and electrochemical energy storage device constraints to ensure the stability of the multi-timescale optimization scheduling model of the wind-solar-storage microgrid. These constraints are specifically expressed by the following formulas: (33) In the formula: for Real-time system load power, (34) In the formula: For new energy Efforts made at all times; For new energy installed capacity, (35) In the formula: , These are the inverter's rated charging power and rated discharging power, respectively. State of charge of the energy storage device; , These represent the upper and lower limits of the state of charge of the energy storage device.

Images