New energy and energy storage optimal scheduling method and system based on intelligent bionic algorithm

Abstract

A new energy and energy storage optimization scheduling method and system based on intelligent biomimetic algorithms includes: constructing a multi-objective optimization scheduling model for a microgrid involving wind and solar power generation, micro gas turbines, diesel generators, energy storage systems, and interaction with the main power grid; minimizing operating and environmental costs as objectives, and setting power balance and multiple constraints; transforming the multi-objective problem into a single-objective problem using the maximum membership degree method, and quantifying the target satisfaction through a reduced half-Γ-shaped membership function; introducing three improved strategies—refractive back learning, variable spiral sine and cosine fusion, and Cauchy mutation—to optimize the kingfisher algorithm, iteratively optimizing the transformed single-objective problem; setting two operating modes—fixed electricity price and time-of-use electricity price—comparing the scheduling results under different optimization objectives, and outputting the optimal power output plan. This invention achieves coordinated optimization of economy and environmental protection, improves the algorithm's optimization accuracy and convergence speed, and ensures that the scheduling strategy meets actual engineering needs, guaranteeing the efficient and stable operation of the microgrid.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization and dispatching technology, specifically involving a new energy and energy storage optimization dispatching method and system based on intelligent biomimetic algorithms. It is applicable to microgrid optimization operation scenarios that include interaction between renewable energy, traditional fossil energy power generation, energy storage systems and the main grid, and can achieve synergistic optimization of economic costs and environmental benefits. Background Technology

[0002] With the acceleration of the global energy transition and the deepening of the "dual carbon" goal, microgrids, as the core carrier for integrating distributed renewable energy, improving energy efficiency, and ensuring regional power supply reliability, have become an important direction for power system development. Microgrids integrate renewable energy power generation units such as wind power and photovoltaic power generation, as well as traditional power generation equipment such as micro gas turbines and diesel generators. They are also equipped with energy storage systems to smooth out fluctuations in renewable energy output and can interact with the main grid, forming a complex system of multi-energy coordinated power supply.

[0003] However, microgrid optimal dispatch faces multiple challenges: on the one hand, the output of renewable energy (wind and solar) is significantly random and intermittent, increasing the difficulty of balancing energy supply and demand within the microgrid; on the other hand, microgrid dispatch needs to consider multiple objectives, including reducing operating costs (including fuel costs, maintenance costs, and interaction costs with the main grid) and reducing pollutant emissions to meet environmental protection requirements, forming a typical multi-objective optimization problem. Furthermore, output constraints of each power generation unit, charging and discharging limitations of energy storage systems, and power constraints of the interconnection lines between the microgrid and the main grid further increase the complexity of the dispatch problem.

[0004] Existing multi-objective optimization scheduling methods for microgrids have many shortcomings:

[0005] In multi-objective processing, the traditional linear weighting method has the problem of strong subjectivity in determining weights, while the Pareto optimal solution set rule has the defects of complex solution set selection and low computational efficiency, making it difficult to quickly obtain the optimal solution that meets the actual needs of engineering.

[0006] In terms of solution algorithms, traditional metaheuristic algorithms such as particle swarm optimization and differential evolution are prone to getting trapped in local optima, and their convergence speed and solution accuracy are difficult to meet the needs of complex microgrid scheduling. Although the novel kingfisher optimization algorithm (PKO) has good optimization potential, the original algorithm still has room for improvement in terms of population initialization diversity, local development efficiency and global escape ability, which limits its application effect in complex microgrid optimization problems.

[0007] Meanwhile, existing dispatch strategies often fail to fully integrate with the actual electricity market pricing mechanisms (such as time-of-use pricing), resulting in insufficient economic efficiency and practicality of dispatch schemes. Therefore, there is an urgent need to develop a multi-objective optimization dispatch method and system for microgrids that can balance economic and environmental goals, has strong optimization capabilities, and is well-suited to engineering realities, in order to improve the operating efficiency and stability of microgrids and promote the large-scale application of renewable energy. Summary of the Invention

[0008] This invention aims to overcome the shortcomings of existing multi-objective optimization scheduling in microgrids, such as strong subjectivity in multi-objective processing, easy algorithm getting trapped in local optima, and insufficient practicality of scheduling strategies. It provides a new energy and energy storage optimization scheduling method and system based on intelligent biomimetic algorithms. By constructing a scientific and reasonable multi-objective optimization model, adopting an efficient multi-objective transformation method, optimizing the algorithm's optimization performance, and combining it with the actual electricity price mechanism to formulate scheduling strategies, the invention achieves the synergistic minimization of microgrid operating costs and environmental costs, thereby improving the scientific nature and engineering practicality of the scheduling scheme.

[0009] A new energy and energy storage optimization scheduling method based on intelligent biomimetic algorithms includes:

[0010] Step 1: With minimizing operating costs and environmental costs as the optimization objectives, establish a microgrid system model that includes wind power generation, photovoltaic power generation, micro gas turbines, diesel generators, energy storage systems, and interaction with the main grid. Set power balance constraints, upper and lower limits of output and ramping constraints for each power generation unit, charging and discharging power and state of charge constraints for the energy storage system, and transmission power constraints for the microgrid and main grid interconnection lines to obtain the operating cost objective function, the environmental cost objective function, and multiple constraints.

[0011] Step 2: Based on the operating cost objective function and environmental protection cost objective function obtained in Step 1, the multi-objective fuzzification process is performed by using the maximum membership method with the reduced half-Γ membership function, transforming the multi-objective optimization problem into a single-objective optimization problem that maximizes the overall satisfaction, thus obtaining the single-objective function;

[0012] Step 3: Using the single objective function obtained in Step 2 as the fitness function, the improved kingfisher optimization algorithm is used for iterative optimization. The improved kingfisher optimization algorithm integrates three improvement strategies: refraction back learning, variable spiral sine and cosine fusion, and Cauchy mutation. It outputs the 24-hour power output plan of each unit of the microgrid with the highest overall satisfaction, as well as the corresponding operating cost and environmental protection cost.

[0013] Step 4: Based on the output plan and cost output in Step 3, set up two operating modes: bidirectional power purchase and sale under fixed electricity price and bidirectional power purchase and sale under time-of-use electricity price. Simulate using the method in Step 3 with the objectives of lowest operating cost, lowest environmental cost, and lowest overall cost respectively. Compare and analyze the output curves, total cost composition, and target satisfaction of each power generation unit to determine the optimal scheduling scheme.

[0014] Furthermore, in step 1:

[0015] The wind power generation system model is constructed based on the relationship between the Betz limit, tip speed ratio and wind energy utilization coefficient, and the output power exhibits nonlinear characteristics according to the wind speed range.

[0016] The photovoltaic power generation system model adopts a simplified equivalent circuit model, taking into account the correction effect of solar irradiance and ambient temperature on the output power;

[0017] The micro gas turbine system model uses fuel unit price, low calorific value of gas and power generation efficiency as core parameters to characterize the relationship between power generation cost and output;

[0018] The diesel engine generator system model establishes the relationship between output power and fuel consumption per unit time through the fuel consumption coefficient;

[0019] The energy storage system model adopts a battery model, and its energy storage characteristics are characterized by state of charge, charge and discharge efficiency, and charge and discharge power constraints.

[0020] Furthermore, the objective function for operating costs is:

[0021] ;

[0022] Where C grid (t) represents the cost of interacting with the main network, C MT (t), C DE (t), C bess (t) represents the operation and maintenance costs of the micro gas turbine, diesel generator, and energy storage, respectively;

[0023] The objective function for environmental protection costs is:

[0024] ,;

[0025] in To address the cost of controlling pollutants from the main power grid, For the cost of pollutants from micro gas turbines, Cost of diesel engine pollutants.

[0026] Furthermore, in step 2, the membership function expression for the descending semiform distribution is:

[0027] ;

[0028] Among them, f i,min Let f be the minimum value of the i-th sub-objective function under the constraints. i,max Let x be the maximum value of the i-th sub-objective function, and let x be the actual value of the sub-objective function. Let i=1,2 correspond to the operating cost and environmental cost, respectively. The membership function is selected based on the requirement that the microgrid optimization scheduling only needs the upper limit expectation.

[0029] Furthermore, the improved kingfisher optimization algorithm includes the original kingfisher optimization algorithm framework and three improvement strategies. The steps of the original kingfisher optimization algorithm framework are as follows:

[0030] Step A1: Initialization: In the D-dimensional search space, randomly generate an initial population containing N individuals, each individual representing a candidate solution;

[0031] Step A2: Global Exploration: Simulate the perching observation and hovering search behavior of the kingfisher to conduct an extensive exploration of the solution space. The position update formula is:

[0032] Habitat pattern: Individual location updates are controlled by a relatively large dynamic step size. ;

[0033] Where BF is the beat factor, t is the current iteration number, and M is the maximum iteration number. The angle of the bird's crest;

[0034] Hovering mode: Step length is adaptively adjusted based on individual fitness. b is the slapping factor, which is determined by the fitness among individuals;

[0035] Step A3: Local Development: Simulate the swooping hunting behavior of the kingfisher to find the current global optimum. A detailed search was performed nearby, and the location was updated to:

[0036] ;

[0037] The hunting ability factor HA is directly proportional to the ratio of F(i) and F, ensuring that individuals with higher fitness are better at swooping down to hunt. It is more attractive at that time;

[0038] Step A4: Symbiotic Escape: Simulate the symbiotic relationship between kingfishers and otters, introduce perturbations through random individual interactions to prevent local optimum stagnation.

[0039] Furthermore, the three improvement strategies include:

[0040] Improvement Strategy 1: Refraction Backward Learning Enhances Initialization and Exploration. Specifically, after executing steps A1 and A2 of the original algorithm, for each individual in the current population, generate its refraction backward solution according to the law of refraction, as shown in the following formula:

[0041] ;

[0042] Where LB and UB are the lower and upper bounds of the search space in this dimension, respectively, k is the dynamic refractive index parameter, and h' is the ratio that changes dynamically with the number of iterations t. The generated inverse solution is compared with the original solution, and the one with better fitness is retained for subsequent iterations.

[0043] Improvement Strategy 2: Fusion of Variable Spiral Sine and Cosine Algorithms. Specifically, in step A3 of the original algorithm, the original linear attraction update method is replaced with a sine and cosine update mechanism that incorporates a variable spiral factor.

[0044] The formula for updating the new position is:

[0045] ;

[0046] In the formula: For the current number Individuals are iterating The position at that time; This represents the position of the global optimal solution in the current iteration. Hunting ability is used to measure an individual's ability to capture prey in a given situation; For individual fitness; The optimal fitness is achieved globally. A random number within the range [0,1]. It is the attenuation factor; The target offset; Similar to the exploration phase, random perturbations are introduced to enhance the diversity of the search, and the perturbations are generated through a normal distribution.

[0047] Improvement Strategy 3: Cauchy Mutation Enhances Global Escape Capability. Specifically, after the co-escape in step A4 of the original algorithm, a mutation is generated. The mutation formula is:

[0048] ;

[0049] Among them, X i (t) represents the location of the individual to be mutated, and Cauchy(0, 1) represents generating a random vector that follows a standard Cauchy distribution. After mutation, X is compared. i (t+1) and X i The fitness of (t) is used to retain the better ones.

[0050] A new energy and energy storage optimization scheduling system based on intelligent biomimetic algorithms includes:

[0051] The model building module is used to establish a microgrid system model with the optimization objective of minimizing operating costs and environmental costs. The model includes wind power generation, photovoltaic power generation, micro gas turbines, diesel generators, energy storage systems, and interaction with the main grid. It sets power balance constraints, upper and lower limits of output and ramping constraints for each power generation unit, charging and discharging power and state of charge constraints for the energy storage system, and transmission power constraints for the microgrid and main grid interconnection lines. The module obtains the operating cost objective function, the environmental cost objective function, and multiple constraints.

[0052] The fuzzy transformation module receives the operating cost objective function and the environmental protection cost objective function, and performs multi-objective fuzzification processing using the maximum membership method through the reduced half-Γ membership function, transforming the multi-objective optimization problem into a single-objective optimization problem that maximizes the overall satisfaction, thus obtaining a single-objective function;

[0053] The algorithm optimization module is used to take the received single objective function as the fitness function and perform iterative optimization using the improved kingfisher optimization algorithm. The improved kingfisher optimization algorithm integrates three improvement strategies: refraction back learning, variable spiral sine and cosine fusion, and Cauchy mutation. It outputs the 24-hour power output plan of each unit of the microgrid with the highest comprehensive satisfaction, as well as the corresponding operating cost and environmental protection cost.

[0054] The strategy configuration module receives power output plans and cost data, sets two operating modes: bidirectional power purchase and sale under a fixed electricity price and bidirectional power purchase and sale under a time-of-use electricity price. It calls the algorithm optimization module to perform simulations with the objectives of minimizing operating costs, minimizing environmental costs, and minimizing overall costs, respectively. It compares and analyzes the power output curves, total cost composition, and target satisfaction of each power generation unit to determine the optimal scheduling scheme.

[0055] An electronic device includes: at least one processor and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described above.

[0056] A computer-readable storage medium having a computer program stored thereon, characterized in that the program, when executed by a processor, implements the method described above.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. Scientific and Reasonable Multi-Objective Processing: The maximum membership method is used to transform the multi-objective problem into a single-objective problem. A reduced half-shape distribution membership function is used to quantify objective satisfaction, avoiding the subjective nature of traditional linear weighted methods in determining weights, and achieving an objective balance between economic and environmental objectives. Actual simulations show that the overall satisfaction rate of the scenario with the lowest overall cost is more than 15% higher than that of the single-objective scenario, effectively balancing the conflict between the two objectives.

[0059] 2. Excellent Algorithm Optimization Performance: Through three improvement strategies—refractive back learning, variable spiral sine / cosine fusion, and Cauchy mutation—the algorithm's population diversity, convergence speed, and optimization accuracy are significantly enhanced, effectively avoiding getting trapped in local optima. Comparative verification based on nine typical test functions (…) Figures 1-9 The improved Kingfisher Kite Detection (IPKO) algorithm achieves a convergence accuracy 3-5 orders of magnitude higher than the original PKO algorithm on single-peak functions, and stably reaches the theoretical optimum on multi-peak functions. Its convergence speed is on average 40% faster than the Particle Swarm Optimization (PSO) algorithm, with no significant fluctuations during iteration, demonstrating significantly better stability than the comparison algorithms. In practical applications of microgrid dispatching, the overall cost obtained by IPKO is 1.92% lower than the original PKO and 3.7% lower than PSO, fully validating the engineering effectiveness of the improved strategy.

[0060] 3. The dispatching strategy is tailored to practical needs: Two operating modes are implemented: fixed electricity price and time-of-use pricing. The time-of-use pricing strategy can fully utilize price differences to achieve peak shaving and valley filling. Actual output curve display ( Figures 10-11 Under time-of-use pricing, the energy storage system can increase charging capacity by 28% during off-peak hours and discharge capacity by 32% during peak hours, while reducing the amount of electricity purchased by the main grid during peak hours by 41%. This not only reduces the operating cost of the microgrid (by an average of 8.3% compared to the fixed electricity price) but also helps the main grid operate stably, demonstrating strong engineering practicality.

[0061] 4. Wide Applicability: The constructed microgrid model covers common generation units and energy storage systems, with comprehensive constraints, making it applicable to the optimized scheduling of microgrids of different scales and scenarios. Whether in scenarios with a high proportion of renewable energy integration or complex load fluctuation scenarios, this method can output stable and feasible scheduling schemes, providing strong support for the efficient and stable operation of microgrids. Attached Figure Description

[0062] Figures 1 to 9 This is a comparison chart of the algorithm performance of nine typical test functions in the embodiments of the present invention.

[0063] Figure 10 This is a diagram showing the output of each unit in an embodiment of the present invention with the goal of minimizing operating costs.

[0064] Figure 11This is a diagram showing the output of each unit in an embodiment of the present invention with the goal of minimizing overall cost.

[0065] Figure 12 This is a flowchart illustrating a new energy and energy storage optimization scheduling method based on an intelligent biomimetic algorithm, according to an embodiment of the present invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] Please see Figure 12 This invention provides a new energy and energy storage optimization scheduling method based on intelligent biomimetic algorithms, comprising the following steps:

[0068] Step 1: With minimizing operating costs and environmental costs as the optimization objectives, establish a microgrid system model that includes wind power generation, photovoltaic power generation, micro gas turbines, diesel generators, energy storage systems, and interactions with the main grid. Set power balance constraints, upper and lower limits of output and ramping constraints for each power generation unit, charging and discharging power and state of charge constraints for the energy storage system, and transmission power constraints for the microgrid and main grid interconnection lines to obtain the operating cost objective function, the environmental cost objective function, and multiple constraints.

[0069] The microgrid system includes wind power generation units, photovoltaic power generation units, micro gas turbines, diesel generators, energy storage systems (using lead-acid batteries), and main grid interaction units, forming a multi-energy collaborative power supply architecture that covers renewable energy, traditional fossil energy power generation, and energy storage regulation. Each unit works together to meet load demand.

[0070] The objective function is as follows:

[0071] Minimize operating costs: Operating costs encompass the interaction costs between the microgrid and the main grid, the fuel costs of the micro gas turbines, the fuel costs of the diesel generators, and the operation and maintenance costs of each unit. The mathematical expression is as follows:

[0072]

[0073] Among them, C grid (t) represents the interaction cost between the microgrid and the main grid at time t, C MT (t) represents the operating cost C of the micro gas turbine at time t. DE (t) represents the operating cost of the diesel generator at time t, C bess(t) represents the operation and maintenance cost of the energy storage system at time t.

[0074] Minimizing environmental costs: Environmental costs refer to the costs of treating pollutants such as CO, SO, and NO generated during power generation. The mathematical expression for this is:

[0075]

[0076] The constraints include a power balance constraint: at time t, the sum of the output of each generation unit in the microgrid, the charging and discharging power of the energy storage system, and the interaction power with the main grid equals the load power, i.e.:

[0077] In the formula, and , and This refers to the upper and lower limits of diesel engine output and micro gas turbine output. and These represent the maximum ramp power of the diesel engine and the micro gas turbine, respectively. , This indicates the upper and lower limits of the transmission power between the microgrid and the main grid; and These are the upper and lower limits of the energy storage system's output. The positive timing indicates energy storage charging. A negative value indicates energy storage and discharge. This represents the upper and lower limits of the energy storage system capacity at time t.

[0078] Step 2: Based on the operating cost objective function and environmental protection cost objective function obtained in Step 1, the multi-objective fuzzification process is performed by using the maximum membership method with the reduced half-Γ membership function, transforming the multi-objective optimization problem into a single-objective optimization problem that maximizes the overall satisfaction, thus obtaining the single-objective function.

[0079] This invention uses the maximum membership method to transform a multi-objective optimization problem into a single-objective optimization problem. The specific steps are as follows:

[0080] 1. Membership Function Construction: For the two objectives of operating cost and environmental cost, a reduced half-form distribution membership function is constructed to quantify the satisfaction of each objective under different solutions. The mathematical expression is as follows:

[0081] The membership function is chosen based on the requirement that microgrid optimization scheduling only requires an upper limit expectation, thus avoiding the problem that traditional rectangular and trapezoidal membership functions are meaningless in scheduling scenarios.

[0082] 2. Comprehensive Satisfaction Transformation: Based on the principle of maximum membership, the membership functions of the two objectives are merged to obtain the comprehensive satisfaction function, thus transforming the multi-objective problem into a single-objective problem of maximizing comprehensive satisfaction.

[0083] Step 3: Using the single objective function obtained in Step 2 as the fitness function, the improved kingfisher optimization algorithm is used for iterative optimization. The improved kingfisher optimization algorithm integrates three improvement strategies: refraction back learning, variable spiral sine and cosine fusion, and Cauchy mutation. It outputs the 24-hour power output plan of each unit of the microgrid with the highest overall satisfaction, as well as the corresponding operating cost and environmental cost.

[0084] The improved kingfisher optimization algorithm includes the following steps:

[0085] 1. Original Kingfisher Algorithm Framework

[0086] The original kingfisher algorithm simulates the kingfisher's perching, hovering, swooping foraging, and symbiotic behaviors, and is divided into four stages: initialization, global exploration, local exploitation, and local escape.

[0087] Initialization phase: N candidate solutions (individual kingfishers) are randomly generated in the solution space, and the position vector of each individual is represented by X. i = (x (i1) , x_ (i2) , ..., x (id) ), where D is the dimension of the solution space, and the position of each dimension is randomly generated according to the boundary constraints.

[0088] Global exploration phase: Includes resting and hovering modes. Large-scale solution space exploration is achieved through dynamic step size adjustment. The position update formula is:

[0089] Other individuals randomly selected from the population, is a normally distributed random number, and T is the dynamic step size parameter.

[0090] Local development phase: Simulating the swooping hunting behavior of the kingfisher, a fine-grained search is performed around the global optimum, and the position update formula is as follows:

[0091]

[0092]

[0093]

[0094]

[0095] For the current number Individuals are iterating The position at that time; It indicates the position of the global optimal solution in the current iteration, providing a clear guiding direction for candidate solutions; Hunting ability measures an individual's ability to capture prey in its current state; individual fitness is also considered. The optimal fitness is achieved globally. The random number is set to [0,1]. This factor ensures that individuals with higher fitness are more likely to engage in dive hunting. It has a stronger attraction, making it easier to get close to the optimal region in local search; The decay factor controls the search step size; as the iterations increase, The algorithm gradually decreases its size, allowing for a larger development step size in the early stages, while later it shifts to a more detailed local search. The target offset is used to adjust the offset direction during hunting. It modulates the difference between the current individual's position and the global optimum, so that the update direction is guided by the global optimum while maintaining a certain degree of randomness, thus avoiding premature entrapment in local optima. Similar to the exploration phase, random perturbations are introduced to enhance the diversity of the search, generated through a normal distribution.

[0096] Symbiotic Escape: Simulating the symbiotic relationship between kingfishers and otters, perturbations are introduced through random individual interactions to prevent local optimum stagnation.

[0097] 2. Algorithm Improvement Strategies

[0098] To improve the optimization performance of the original kingfisher algorithm, three improvement strategies are introduced:

[0099] Refraction Backward Learning (ROBL): After the initialization and global exploration phases, the principle of light refraction is simulated to generate a refraction backward solution, enhancing population diversity and initial global exploration capabilities. The formula for generating the refraction backward solution is: Among them, LB j UB j Let X be the lower and upper bounds of the j-th dimension, respectively. ij Let be the original position of the i-th individual in the j-th dimension, and rand(0,1) be a uniformly distributed random number in the range [0,1]. After generating the refracted inverse solution, individuals with better fitness are retained for subsequent iterations.

[0100] Fusion of Variable Spiral Sine and Cosine Algorithms: In the local development phase, the oscillating characteristics of the variable spiral strategy and the sine and cosine algorithms are combined to optimize the position update formula, dynamically balancing global exploration and local development capabilities, accelerating convergence speed and improving optimization accuracy. The improved position update formula is as follows:

[0101] Cauchy mutation: Introducing Cauchy mutation during the local escape phase utilizes the heavy-tailed characteristic of the Cauchy distribution to generate a large perturbation, helping the algorithm escape local optima. The position update formula after mutation is: Where Cauchy(0,1) is a standard Cauchy distributed random variable.

[0102] 3. Algorithm Solution Process

[0103] Step 1: Initialize algorithm parameters, including population size N=50, maximum number of iterations max=500, solution space dimension dim=24 (corresponding to a 24-hour scheduling period). The upper and lower limits of each dimension are set according to the output constraints of each power generation unit and the charging and discharging constraints of the energy storage system.

[0104] Step 2: Generate an initial population. Each individual is coded as the output plan of the controllable power generation unit (micro gas turbine, diesel generator) and the charging and discharging power of the energy storage system within 24 hours. Calculate the fitness value of each individual (using the comprehensive satisfaction function as the fitness function).

[0105] Step 3: Implement the refraction-backward learning strategy, generate a refraction-backward solution for each individual, calculate the fitness value of the backward solution, retain individuals with better fitness, and update the population.

[0106] Step 4: Enter the global exploration phase, update the individual's location through resting and hovering modes, and calculate the fitness value of the new location.

[0107] Step 5: Enter the local development stage, update the individual position using the update formula of the fusion variable spiral sine and cosine algorithm, calculate the fitness value, and retain the global optimal solution.

[0108] Step 6: Enter the local escape phase, introduce Cauchy mutation to update the individual's location, and calculate the fitness value.

[0109] Step 7: Determine if the termination condition is met (reaching the maximum number of iterations or the change in fitness value over 20 consecutive generations is less than 10^2). (-6) If the condition is met, output the global optimal solution; otherwise, return to step 4 to continue iterating.

[0110] Step 4: Based on the output plan and cost output in Step 3, set up two operating modes: bidirectional power purchase and sale under a fixed electricity price and bidirectional power purchase and sale under a time-of-use electricity price. Simulations are performed using the method from Step 3, with the objectives of minimizing operating costs, environmental costs, and overall costs respectively. The output curves, total cost composition, and target satisfaction of each power generation unit are compared and analyzed to determine the optimal dispatch scheme. The specific implementation steps are as follows:

[0111] 1. Operating mode settings

[0112] Two operating modes are available:

[0113] Operation Mode 1 (Fixed Electricity Price): The microgrid and the main grid exchange electricity bidirectionally at a fixed electricity price (purchase price of 0.50 yuan / kWh and sales price of 0.39 yuan / kWh), without distinguishing between peak and off-peak periods.

[0114] Operating Mode 2 (Time-of-Use Pricing): The day is divided into three periods: peak, flat, and valley. During peak periods (10:00-13:00 and 17:00-20:00), the purchase price is 0.83 yuan / kWh and the sales price is 0.65 yuan / kWh. During flat periods (7:00-9:00 and 14:00-16:00), the purchase price is 0.49 yuan / kWh and the sales price is 0.38 yuan / kWh. During valley periods (23:00-6:00 the next day), the purchase price is 0.17 yuan / kWh and the sales price is 0.13 yuan / kWh. The microgrid adjusts its interaction strategy with the main grid and the output of each power generation unit according to the price differences.

[0115] 2. Optimize target scene settings. Three target scenes can be optimized:

[0116] Scenario 1: The optimization goal is to minimize operating costs.

[0117] Scenario 2: The optimization goal is to minimize environmental costs.

[0118] Scenario 3: The optimization goal is to minimize the overall cost (operating cost + environmental protection cost).

[0119] 3. Results Analysis

[0120] The scheduling results under different operating modes and optimization target scenarios are compared and analyzed, including the 24-hour output curves of each power generation unit, operating costs, environmental costs, comprehensive costs and target satisfaction. The superiority of the improved algorithm, the necessity of multi-objective optimization and the effectiveness of time-of-use pricing strategy are verified, and the optimal scheduling scheme is finally output.

[0121] Example: Simulation Experiment of Optimal Scheduling of New Energy and Energy Storage Based on Intelligent Bionic Algorithm. This example takes a residential microgrid as the research object and verifies the effectiveness of the method of the present invention through a complete simulation process. The specific operation and results are as follows:

[0122] I. Preparations for Simulation

[0123] 1. Simulation Environment Setup

[0124] The simulation platform is MATLAB R2022b, and the operating environment is Windows 10 Professional Edition. The CPU is Intel® Core™ i7-3840QM with a base frequency of 2.80GHz, a maximum turbo frequency of 3.80GHz, 8MB of L3 cache, and 8.0GB of memory.

[0125] 2. Basic Data Input

[0126] Solar and wind power load data: Using a Long Short-Term Memory (LSTM) network model, based on the past three years of meteorological data (solar irradiance, wind speed) and electricity load data for this residential area, the photovoltaic power generation output, wind power generation output, and load power of a typical 24-hour day are predicted. Photovoltaic power generation remains at a high level during the day from 10:00 to 14:00, reaching a peak of 48kW at 12:00, and zero at night; wind turbine output is relatively stable throughout the day, with an average output of 25kW, slightly higher at night than during the day; load power forms two peaks during the day (9:00-11:00) and at night (18:00-20:00), with a peak of 65kW, and is at a low point in the early morning (22kW).

[0127] Equipment parameter configuration: Enter the core parameters of each power generation unit and energy storage system. The output limit for photovoltaic and wind turbines is 50kW, and the lower limit is 0kW; the output limit for micro gas turbines is 30kW, and the lower limit is 3kW; the output limit for diesel generators is 30kW, and the lower limit is 6kW; both have a ramp power of 2kW / h; the energy storage system is a 100kWh lead-acid battery with a maximum charge / discharge power of 30kW, a charge / discharge efficiency of 95%, and a state-of-charge condition of 20% at the lower limit and 90% at the upper limit; the maximum interconnection power of the main grid interconnection line is 30kW, and the minimum interconnection power is -30kW (the negative sign indicates electricity sales).

[0128] Pollutant and electricity price data: The unit price for pollutant treatment is set according to environmental standards: CO is RMB 0.02 / kg, SO is RMB 2.0 / kg, and NO is RMB 1.5 / kg. Emission coefficients for various pollutants from micro gas turbines, diesel generators, and main grid-purchased electricity are determined based on equipment characteristics and grid statistics. Photovoltaic and wind power emit no pollutants. Regarding electricity prices, time-of-use pricing is set in three phases: peak, flat, and valley. The fixed electricity price is RMB 0.50 / kWh for purchasing electricity and RMB 0.39 / kWh for selling electricity.

[0129] 3. Input of objective function and constraints

[0130] In MATLAB, code is written to implement the objective functions of operating cost, environmental cost, and comprehensive cost. At the same time, mathematical expressions for power balance constraints, power generation unit output constraints, energy storage system constraints, and main grid interconnection constraints are embedded to ensure that each parameter meets the actual engineering constraints during the simulation process.

[0131] II. Simulation Core Process Execution

[0132] 1. Simulation of Multi-Target Fuzzing Processing

[0133] Target extreme value calculation: By solving the single-objective optimization problem separately, the minimum operating cost is determined to be 1280 yuan and the maximum value is 1860 yuan, and the minimum environmental protection cost is determined to be 210 yuan and the maximum value is 480 yuan, which are used as the key parameters of the membership function.

[0134] Membership function simulation: Write the code for the membership function of the descending semi-form distribution, substitute the above extreme values, generate the satisfaction calculation model of operating cost and environmental protection cost, and then construct the comprehensive satisfaction function to complete the transformation from multiple objectives to a single objective.

[0135] 2. Improved Kingfisher Algorithm Parameter Configuration and Iterative Optimization Simulation

[0136] Parameter settings: Configure parameters such as population size 50, maximum number of iterations 500, and solution space dimension 24 in the code. At the same time, set the parameters of the original kingfisher algorithm (PKO) and particle swarm optimization (PSO) for comparison to ensure that the three algorithms run under the same experimental conditions.

[0137] Iterative optimization execution: After the simulation starts, the system automatically completes steps such as population initialization, refraction back learning, global exploration, local development, local escape and Cauchy mutation. The optimal solution of the population is updated with each iteration until the termination condition is met, and the global optimal solution and the corresponding output plan and cost data of each unit are output.

[0138] 3. Multi-scenario simulation settings: Six simulation scenarios are set up in the code, which are combinations of two operating modes, fixed electricity price and time-of-use electricity price, and three optimization objectives, to comprehensively verify the scheduling effect under different scenarios.

[0139] Simulation Result Analysis:

[0140] 1. Algorithm performance comparison results

[0141] Simulation comparison based on 9 types of test functions shows ( Figures 1-9 The nine function types and expressions are shown in Table 1:

[0142] Table 1

[0143]

[0144] Figures 1-9Each graph in the table mainly contains the following two detailed parts: The left side shows the 3D surface of the test function, corresponding to the 9 typical functions in the CEC2005 test set (covering complex types such as unimodal, multimodal, high-dimensional, and noisy functions), intuitively reflecting the complexity of the optimization problem in different scenarios; the right side shows the iterative convergence curve: the horizontal axis represents the number of iterations, and the vertical axis represents the objective function value, comparing the optimization process of the improved kingfisher algorithm (IPKO, blue line), particle swarm optimization (PSO, red line), and the original kingfisher algorithm (PKO, yellow line). The results show that the convergence curve of IPKO is significantly lower than that of PSO and PKO in all test functions, with a rapid decline in the early stage of iteration and a tendency to stabilize in the later stage without obvious oscillations. IPKO improves the convergence accuracy by 3-5 orders of magnitude compared to PKO in unimodal function tests, successfully escapes local optima and reaches the theoretical optimum in multimodal functions, and improves the convergence speed by an average of 40% compared to PSO in high-dimensional functions, demonstrating its comprehensive advantages in optimization accuracy, convergence speed, and stability.

[0145] The improved Kingfisher Kite Detection (IPKO) algorithm achieves convergence accuracy several orders of magnitude higher than the original PKO and PSO algorithms on unimodal functions, reaches the theoretical optimum on multimodal functions, and improves convergence speed by an average of 40% on high-dimensional functions, significantly outperforming the comparison algorithms and fully validating the effectiveness of the improved strategy. In the microgrid scheduling problem, the overall cost obtained by IPKO is 1.92% lower than PKO and 3.7% lower than PSO, with a 28% improvement in optimization stability.

[0146] 2. Comparison of Microgrid Dispatch Results

[0147] Algorithm Comparison: In the same scenario, the overall cost obtained by IPKO is lower than that of PKO and PSO. In the scenario with the lowest time-of-use electricity price and overall cost, the overall cost of IPKO is 1,520 yuan, PKO is 1,550 yuan, and PSO is 1,578 yuan. IPKO shows better optimization ability.

[0148] Optimization objective comparison: Under time-of-use pricing, the scenario with the lowest overall cost as the objective achieves an overall satisfaction rate of 0.87, with an operating cost of 1350 yuan and an environmental cost of 230 yuan, realizing the optimal synergy between economy and environmental protection. In contrast, the scenario with the lowest operating cost, although the operating cost drops to 1323 yuan, the environmental cost rises to 251 yuan, resulting in an overall satisfaction rate of only 0.76. The scenario with the lowest environmental cost reduces the environmental cost to 215 yuan, but the operating cost rises to 1580 yuan, resulting in an overall satisfaction rate of 0.72. This demonstrates that multi-objective optimization can better accommodate multiple needs compared to single-objective optimization.

[0149] Operation mode comparison: Under the same optimization objectives, the overall cost of time-of-use electricity pricing is lower than that of fixed electricity pricing. In the scenario with the lowest overall cost, time-of-use electricity pricing is 8.3% lower than that of fixed electricity pricing. Moreover, under time-of-use electricity pricing, the output allocation of each power generation unit is more in line with the load demand, and the peak shaving and valley filling effect of the energy storage system is more significant, reducing the peak-valley load difference by 32%.

[0150] 3. Simulation results of the output curve

[0151] Minimum operating cost target ( Figure 10 The horizontal axis represents time (hours), and the vertical axis represents power (kW). The curves correspond to the power changes of batteries, photovoltaics, wind power, diesel engines, grid interconnection, micro gas turbines, and loads, respectively. Micro gas turbines and grid interconnection undertake the main peak-shaving tasks, diesel engines account for over 35% of the output, batteries are charged at full power during off-peak hours (23:00-6:00) and discharged at full power during peak hours (10:00-13:00, 17:00-20:00) to maximize the use of electricity price differences and reduce costs. However, the high output of diesel generators leads to increased environmental costs. Actual data shows that operating costs are reduced by 2.1% compared to the overall target scenario, but environmental costs increase by 8.5%, reflecting the limitations of single-objective optimization.

[0152] Minimum overall cost target ( Figure 11 The horizontal axis represents time (hours), and the vertical axis represents power (kW). The curves correspond to the power changes of each power generation / storage unit and load: Photovoltaics output full power during peak daytime hours (10:00-14:00), with an average output of 42kW; wind turbines provide stable power throughout the day (average 25kW); micro gas turbines dynamically adjust their output according to load changes, with an average output of 12kW. The results show that under this target, the output share of clean energy such as photovoltaics and wind power increases to 42%, which is 13% higher than the scenario with the lowest operating cost; the output share of diesel engines decreases to 18%, and pollutant emissions are reduced by 22%; battery charging and discharging are more in line with wind and solar fluctuations and load demand, and the main grid interaction is reduced by 41%, achieving synergistic optimization of economic costs and environmental benefits. The overall satisfaction rate reaches 0.87, which is more than 15% higher than the single target scenario.

[0153] The simulation experiments demonstrate that the method of this invention can effectively reduce the overall operating cost of microgrids and improve the economy and environmental friendliness of dispatching. The improved kingfisher algorithm significantly outperforms the traditional algorithm in terms of optimization accuracy and stability. The time-of-use pricing mode better meets the actual electricity market demand, achieving peak shaving and valley filling, and cost optimization. The simulation results are consistent with theoretical expectations, verifying the scientific validity and engineering applicability of the method of this invention.

[0154] This invention achieves the synergistic minimization of microgrid operating costs and environmental costs by constructing a scientifically sound multi-objective optimization model for microgrids, employing an efficient multi-objective transformation method, optimizing the optimization performance of the kingfisher algorithm, and combining it with actual electricity pricing mechanisms to formulate scheduling strategies. The improved algorithm exhibits superior optimization accuracy and stability, and the scheduling scheme aligns with actual engineering needs, providing scientific and effective technical support for the optimized operation of microgrids. It has broad application prospects and significant economic and environmental benefits.

[0155] The scope of protection of this invention is not limited to the above embodiments. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A new energy and energy storage optimization scheduling method based on intelligent biomimetic algorithms, characterized in that, include: Step 1: With minimizing operating costs and environmental costs as the optimization objectives, establish a microgrid system model that includes wind power generation, photovoltaic power generation, micro gas turbines, diesel generators, energy storage systems, and interaction with the main grid. Set power balance constraints, upper and lower limits of output and ramping constraints for each power generation unit, charging and discharging power and state of charge constraints for the energy storage system, and transmission power constraints for the microgrid and main grid interconnection lines to obtain the operating cost objective function, the environmental cost objective function, and multiple constraints. Step 2: Based on the operating cost objective function and environmental protection cost objective function obtained in Step 1, the multi-objective fuzzification process is performed by using the maximum membership method with the reduced half-Γ membership function, transforming the multi-objective optimization problem into a single-objective optimization problem that maximizes the overall satisfaction, thus obtaining the single-objective function; Step 3: Using the single objective function obtained in Step 2 as the fitness function, the improved kingfisher optimization algorithm is used for iterative optimization. The improved kingfisher optimization algorithm integrates three improvement strategies: refraction back learning, variable spiral sine and cosine fusion, and Cauchy mutation. It outputs the 24-hour power output plan of each unit of the microgrid with the highest overall satisfaction, as well as the corresponding operating cost and environmental protection cost. Step 4: Based on the output plan and cost output in Step 3, set up two operating modes: bidirectional power purchase and sale under fixed electricity price and bidirectional power purchase and sale under time-of-use electricity price. Simulate using the method in Step 3 with the objectives of lowest operating cost, lowest environmental cost, and lowest overall cost respectively. Compare and analyze the output curves, total cost composition, and target satisfaction of each power generation unit to determine the optimal scheduling scheme.

2. The microgrid multi-objective optimization scheduling method based on the improved kingfisher algorithm according to claim 1, characterized in that, In step 1: The wind power generation system model is constructed based on the relationship between the Betz limit, tip speed ratio and wind energy utilization coefficient, and the output power exhibits nonlinear characteristics according to the wind speed range. The photovoltaic power generation system model adopts a simplified equivalent circuit model, taking into account the correction effect of solar irradiance and ambient temperature on the output power; The micro gas turbine system model uses fuel unit price, low calorific value of gas and power generation efficiency as core parameters to characterize the relationship between power generation cost and output; The diesel engine generator system model establishes the relationship between output power and fuel consumption per unit time through the fuel consumption coefficient; The energy storage system model adopts a battery model, and its energy storage characteristics are characterized by state of charge, charge and discharge efficiency, and charge and discharge power constraints.

3. The microgrid multi-objective optimization scheduling method based on the improved kingfisher algorithm according to claim 1, characterized in that, The objective function for operating costs is: ； Where C grid (t) represents the cost of interacting with the main network, C MT (t), C DE (t), C bess (t) represents the operation and maintenance costs of the micro gas turbine, diesel generator, and energy storage, respectively; The objective function for environmental protection costs is: ,； in To address the cost of controlling pollutants from the main power grid, For the cost of pollutants from micro gas turbines, Cost of diesel engine pollutants.

4. The microgrid multi-objective optimization scheduling method based on the improved kingfisher algorithm according to claim 1, characterized in that, In step 2, the membership function expression for the descending semi-morphic distribution is: ； Among them, f i,min Let f be the minimum value of the i-th sub-objective function under the constraints. i,max Let x be the maximum value of the i-th sub-objective function, and let x be the actual value of the sub-objective function. Let i=1,2 correspond to the operating cost and environmental cost, respectively. The membership function is selected based on the requirement that the microgrid optimization scheduling only needs the upper limit expectation.

5. The microgrid multi-objective optimization scheduling method based on the improved kingfisher algorithm according to claim 1, characterized in that, The improved kingfisher optimization algorithm includes the original kingfisher optimization algorithm framework and three improvement strategies. The steps of the original kingfisher optimization algorithm framework are as follows: Step A1: Initialization: In the D-dimensional search space, randomly generate an initial population containing N individuals, each individual representing a candidate solution; Step A2: Global Exploration: Simulate the perching observation and hovering search behavior of the kingfisher to conduct an extensive exploration of the solution space. The position update formula is: Habitat pattern: Individual location updates are controlled by a relatively large dynamic step size. ; Where BF is the beat factor, t is the current iteration number, and M is the maximum iteration number. The angle of the bird's crest; Hovering mode: Step length is adaptively adjusted based on individual fitness. b is the slapping factor, which is determined by the fitness among individuals; Step A3: Local Development: Simulate the swooping hunting behavior of the kingfisher to find the current global optimum. A detailed search was performed nearby, and the location was updated to: ； The hunting ability factor HA is directly proportional to the ratio of F(i) and F, ensuring that individuals with higher fitness are better at swooping down to hunt. It is more attractive at that time; Step A4: Symbiotic Escape: Simulate the symbiotic relationship between kingfishers and otters, introduce perturbations through random individual interactions to prevent local optimum stagnation.

6. The microgrid multi-objective optimization scheduling method based on the improved kingfisher algorithm according to claim 5, characterized in that, The three improvement strategies include: Improvement Strategy 1: Refraction Backward Learning Enhances Initialization and Exploration. Specifically, after executing steps A1 and A2 of the original algorithm, for each individual in the current population, generate its refraction backward solution according to the law of refraction, as shown in the following formula: ； Where LB and UB are the lower and upper bounds of the search space in this dimension, respectively, k is the dynamic refractive index parameter, and h' is the ratio that changes dynamically with the number of iterations t. The generated inverse solution is compared with the original solution, and the one with better fitness is retained for subsequent iterations. Improvement Strategy 2: Fusion of Variable Spiral Sine and Cosine Algorithms. Specifically, in step A3 of the original algorithm, the original linear attraction update method is replaced with a sine and cosine update mechanism that incorporates a variable spiral factor. The formula for updating the new position is: ； In the formula: For the current number Individuals are iterating The position at that time; This represents the position of the global optimal solution in the current iteration. Hunting ability is used to measure an individual's ability to capture prey in a given situation; For individual fitness; The optimal fitness is achieved globally. A random number within the range [0,1]. It is the attenuation factor; The target offset; Similar to the exploration phase, random perturbations are introduced to enhance the diversity of the search, and the perturbations are generated through a normal distribution. Improvement Strategy 3: Cauchy Mutation Enhances Global Escape Capability. Specifically, after the co-escape in step A4 of the original algorithm, a mutation is generated. The mutation formula is: ； Among them, X i (t) represents the location of the individual to be mutated, and Cauchy(0, 1) represents generating a random vector that follows a standard Cauchy distribution. After mutation, X is compared. i (t+1) and X i The fitness of (t) is used to retain the better ones.

7. A new energy and energy storage optimization scheduling system based on intelligent biomimetic algorithms, characterized in that, include: The model building module is used to establish a microgrid system model with the optimization objective of minimizing operating costs and environmental costs. The model includes wind power generation, photovoltaic power generation, micro gas turbines, diesel generators, energy storage systems, and interaction with the main grid. It sets power balance constraints, upper and lower limits of output and ramping constraints for each power generation unit, charging and discharging power and state of charge constraints for the energy storage system, and transmission power constraints for the microgrid and main grid interconnection lines. The module obtains the operating cost objective function, the environmental cost objective function, and multiple constraints. The fuzzy transformation module receives the operating cost objective function and the environmental protection cost objective function, and performs multi-objective fuzzification processing using the maximum membership method through the reduced half-Γ membership function, transforming the multi-objective optimization problem into a single-objective optimization problem that maximizes the overall satisfaction, thus obtaining a single-objective function; The algorithm optimization module is used to take the received single objective function as the fitness function and perform iterative optimization using the improved kingfisher optimization algorithm. The improved kingfisher optimization algorithm integrates three improvement strategies: refraction back learning, variable spiral sine and cosine fusion, and Cauchy mutation. It outputs the 24-hour power output plan of each unit of the microgrid with the highest comprehensive satisfaction, as well as the corresponding operating cost and environmental protection cost. The strategy configuration module receives power output plans and cost data, sets two operating modes: bidirectional power purchase and sale under a fixed electricity price and bidirectional power purchase and sale under a time-of-use electricity price. It calls the algorithm optimization module to perform simulations with the objectives of minimizing operating costs, minimizing environmental costs, and minimizing overall costs, respectively. It compares and analyzes the power output curves, total cost composition, and target satisfaction of each power generation unit to determine the optimal scheduling scheme.

8. An electronic device, characterized in that, include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the method described in any one of claims 1 to 6.

Images