Novel power distribution network two-stage collaborative optimization method, device and equipment considering economical efficiency and operation risk

Abstract

The invention discloses a novel power distribution network two-stage collaborative optimization method, device and equipment considering economical efficiency and operation risks, and relates to the technical field of power system optimization. The method comprises the following steps: acquiring power distribution network parameters and prediction data; constructing a day-ahead-day two-stage optimization operation framework and constraint conditions; establishing a day-ahead stage optimization model taking system operation cost minimization as a target; establishing an intra-day stage optimization model with the aim of minimizing the comprehensive operation risk of the power distribution network; solving the two-stage optimization model by adopting a novel global optimization algorithm; and formulating a power distribution network optimization operation strategy according to a solving result. The method effectively solves the problems that an existing method is difficult to collaboratively optimize economical efficiency and safety, and an optimization algorithm is prone to falling into local optimum, an efficient and reliable optimization tool is provided for the power distribution network with high-proportion new energy access, and the system operation risk is remarkably reduced.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization technology, specifically relating to a novel two-stage collaborative optimization method, device, and equipment for distribution networks that takes into account both economic efficiency and operational risks. Background Technology

[0002] With the large-scale integration of distributed power sources such as wind and solar power into the distribution network, the randomness and volatility of their output pose a severe challenge to the operational safety of the distribution network. Existing methods for optimizing distribution network operation have fundamental limitations in addressing this challenge, primarily at two levels:

[0003] On the one hand, at the optimization framework level, the objectives are singular and the coordination mechanism is lacking. Most existing distribution network optimization methods still focus solely on economic efficiency, such as minimizing operating costs, without considering operational risks like voltage exceedances and line overloads as core optimization objectives. In scenarios with high-proportion renewable energy integration, these methods struggle to achieve an effective balance between economy and safety amidst complex uncertainties, easily leading to theoretically economical optimization schemes that lack sufficient safety margins and pose high risks in actual operation. Although some studies have adopted a day-ahead / intraday two-stage framework, the intraday stage primarily focuses on economic fine-tuning, lacking an independent optimization stage with a clear objective of "minimizing operational risks," resulting in the system's inability to proactively correct for real-time fluctuations.

[0004] On the other hand, at the optimization algorithm level, existing metaheuristic algorithms have inherent flaws. Core algorithms used to solve this complex optimization problem, such as particle swarm optimization and genetic algorithms, generally suffer from insufficient convergence accuracy, susceptibility to local optima, and an imbalance between global exploration and local exploitation capabilities when dealing with such high-dimensional, nonlinear, and multi-constraint distribution network optimization models. The fundamental reason is that the population search mechanism of these algorithms is not strongly guided and easily loses its way in the complex solution space; at the same time, their diversity preservation strategies are insufficient, leading to premature homogenization of the population, making it difficult to continuously and efficiently find high-quality solutions. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a novel two-stage collaborative optimization method, apparatus, and equipment for power distribution networks that considers both economic efficiency and operational risks. The aim is to solve the technical problems in existing power distribution network optimization methods, such as the difficulty in balancing economic efficiency and safety due to structural defects in the optimization framework, and the poor convergence performance and localized optimization results caused by inherent defects in the guidance and diversity of optimization algorithms.

[0006] One embodiment of the present invention provides a novel two-stage collaborative optimization method for distribution networks that takes into account both economic efficiency and operational risks, including:

[0007] Step 1: Obtain distribution network parameters, predicted output of distributed generation, and load forecast data;

[0008] Step 2: Construct a two-stage optimization operation framework and system constraints for the distribution network, covering both day-ahead and intraday phases;

[0009] Step 3: Establish a day-ahead stage optimization model with the goal of minimizing system operating costs;

[0010] Step 4: Establish an intraday phase optimization model with the goal of minimizing the operational risks of the distribution network;

[0011] Step 5: Solve the two-stage optimization model using a novel global optimization algorithm;

[0012] Step 6: Develop a power distribution network optimization operation strategy based on the solution results.

[0013] Furthermore, the solutions to both the day-ahead optimization model and the intraday optimization model described in step 2 must satisfy the system operation constraints of the distribution network. These system operation constraints generally fall into two categories: general system operation constraints and system safety operation boundaries.

[0014] Under the two-stage optimization framework, the strategies for applying the aforementioned constraints differ: In the day-ahead stage, optimization must simultaneously satisfy both the general constraints for all system operations and the system's safe operating boundary. In the intra-day stage, optimization must satisfy the general constraints for all system operations, while transforming the core elements of the system's safe operating boundary into quantifiable risk indicators, which are then minimized as independent optimization objectives. This achieves a strategic shift from "hard constraint satisfaction" to "soft objective optimization," proactively managing the security risks arising from uncertainty.

[0015] Furthermore, the objective function of the day-one optimization model in step 3 is:

[0016]

[0017] In the formula, For traditional unit operating costs; Costs associated with reduced energy storage lifespan; Cost compensation for interruptible loads; Costs associated with interaction with the upstream power grid.

[0018] Furthermore, the objective function of the intraday stage optimization model in step 4 is:

[0019]

[0020] In the formula, , , These are the indicators for node voltage over-limit risk, branch power over-limit risk, and load shedding risk. , , These are the weighting coefficients.

[0021] Furthermore, the novel global optimization algorithm in step 5 is specifically designed to address the high-dimensionality, nonlinearity, and multi-constraint characteristics of distribution network optimization models. It overcomes the shortcomings of traditional algorithms, such as weak guidance and poor diversity, through the synergistic effect of four core mechanisms. The algorithm includes the following steps:

[0022] Step 5.1: Within the search space, randomly generate a set of initial solutions based on the upper and lower bounds of the variables to form the search proxy set of the algorithm, in order to start the optimization process;

[0023] Step 5.2: Employ a residual-guided update mechanism to construct a guiding vector using elite individual information, thereby improving the accuracy of the search direction and the global exploration capability;

[0024] Step 5.3: Implement a double random reference update mechanism, introduce random reference individuals and gradient information in the exploration phase, and explore and develop dynamic balancing algorithms;

[0025] Step 5.4: Implement a weighted multi-path update mechanism. Design a multi-path update strategy during the development phase to enhance the diversity of local searches;

[0026] Step 5.5: Introduce the Lévy flight perturbation mechanism, and execute the Lévy flight perturbation according to the probability of dynamic triggering, so that the algorithm can effectively escape the local optimum region;

[0027] Step 5.6: Determine whether the algorithm has reached the maximum number of iterations or convergence accuracy. If the conditions are met, output the optimal solution; otherwise, return to step 5.2 to continue iterating.

[0028] Another embodiment of the present invention provides a novel two-stage collaborative optimization device for power distribution networks that takes into account both economic efficiency and operational risks, comprising:

[0029] The parameter acquisition module is used to acquire distribution network parameters, distributed generation predicted output, and load forecast data.

[0030] The framework construction module is used to build the day-ahead and intraday two-stage optimization operation framework and constraints for the distribution network.

[0031] The day-ahead model building module is used to build the day-ahead phase optimization model;

[0032] The intraday model building module is used to build intraday phase optimization models.

[0033] The optimization solution module is used to solve the model using a novel global optimization algorithm;

[0034] The control and execution module is used to formulate optimal operation strategies for the distribution network based on the solution results and to control the distribution network.

[0035] Another embodiment of the present invention provides a terminal device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the method described above.

[0036] Another embodiment of the present invention provides a storage medium including a stored computer program, wherein the computer program, when running, controls the device where the storage medium is located to perform the method described above.

[0037] The embodiments of the present invention have the following beneficial effects:

[0038] (1) By constructing a day-to-day two-stage optimization framework with “economic efficiency” and “risk minimization” as clear independent objectives, structural synergistic optimization of operational economy and safety is achieved, overcoming the shortcomings of the existing two-stage framework that only focuses on economic fine-tuning in the day stage and ignores forward-looking safety correction, which is more in line with the actual dispatching needs of the power system under high proportion of new energy access.

[0039] (2) The proposed novel global optimization algorithm significantly improves the search guidance, population diversity and local optimum escape ability of the algorithm in the solution space of complex distribution networks through the synergistic effect of four core mechanisms: residual guidance, double random reference, weighted multipath and Levy flight, thereby obtaining higher quality and more stable optimization solutions.

[0040] (3) It provides a complete, efficient and reliable optimization decision-making tool for distribution networks with a high proportion of distributed power sources. It not only reduces the system operating cost, but more importantly, it effectively ensures the safe and stable operation of the power grid by actively minimizing the operating risks. Attached Figure Description

[0041] Figure 1 This is a flowchart of a novel two-stage collaborative optimization method for power distribution networks that takes into account both economic efficiency and operational risks, provided by an embodiment of the present invention.

[0042] Figure 2 This is a schematic diagram of the framework for a novel two-stage collaborative optimization of a power distribution network provided in an embodiment of the present invention;

[0043] Figure 3 This is a flowchart of a novel global optimization algorithm provided in an embodiment of the present invention;

[0044] Figure 4This is a schematic diagram of the structure of a novel two-stage collaborative optimization device for power distribution networks provided in an embodiment of the present invention;

[0045] Figure 5 This is a schematic diagram of a terminal device and a computer-readable storage medium provided in an embodiment of the present invention; Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0047] Example 1: A novel two-stage collaborative optimization method for power distribution networks that considers both economic efficiency and operational risks, comprising the following steps:

[0048] Step 1: Obtain distribution network parameters, predicted output of distributed generation sources, and load forecast data. Distribution network parameters include line resistance, reactance, node voltage limits, and line current limits. The predicted power of distributed generation sources is generated based on historical data and takes into account uncertainty distribution.

[0049] Step 2: Construct a two-stage optimization operation framework and system constraints for the distribution network, from day-ahead to intraday.

[0050] like Figure 2 As shown, the two-stage optimization framework includes a day-ahead optimization layer and an intraday optimization layer. The day-ahead optimization layer formulates a basic operational plan based on medium- and long-term forecast data, while the intraday optimization layer refines the operational plan based on short-term forecast data.

[0051] The system constraints that the two-stage optimization model needs to satisfy include the following categories. In the day-ahead optimization stage, the optimization needs to simultaneously satisfy the general system operation constraints (1)-(4) and the system safety operation constraints (5). In the intraday optimization stage, the optimization needs to satisfy the general system operation constraints (1)-(4).

[0052] (1) Power flow balance constraints in distribution networks:

[0053] In the formula, and They are nodes exist Net injected active and reactive power at any given time. and They are nodes exist The active and reactive power of the energy storage system at any given time; and They are nodes exist The active and reactive power of photovoltaic power generation at any given time; and They are nodes exist Active and reactive power of the load at any given time; For nodes exist The reactive power of the reactive power compensation device at any given time.

[0054] (2) Unit output limitations:

[0055]

[0056] In the formula, To contribute to the unit, and These represent the lower and upper limits of its output, respectively.

[0057] (3) Energy storage system operation constraints:

[0058]

[0059] In the formula, Let t be the energy level of the energy storage device. , The charging and discharging efficiency of energy storage devices; , For energy storage The charging and discharging status at any given time; , for The charging and discharging power of time-segmented energy storage; This represents the maximum permissible charging / discharging power of the energy storage device. , For energy storage charging and discharging efficiency; The duration of each scheduling period.

[0060] (4) Interruptible load operation constraints:

[0061]

[0062] In the formula, Let be the interrupt power of node j at time t. The maximum number of interrupts allowed.

[0063] (5) System safety operation constraints:

[0064] In the formula, , and These are the maximum calculated values ​​for the node voltage over-limit risk, branch power over-limit risk, and load shedding risk indicators, respectively.

[0065] Step 3: Establish a day-ahead stage optimization model with the objective function of minimizing system operating costs.

[0066] The objective function of the current stage optimization model is:

[0067]

[0068] In the formula, For traditional unit operating costs; Costs associated with reduced energy storage lifespan; Cost compensation for interruptible loads; Costs associated with interaction with the upstream power grid.

[0069] The specific expression for the operating cost of traditional generating units is as follows:

[0070]

[0071] In the formula, The total number of time periods in the scheduling cycle. The total number of gas turbines. This represents the total number of diesel generator sets. , Gas turbines and diesel generator sets, respectively, during the time period Those who have made meritorious contributions; and This represents the cost coefficient for gas turbines and diesel generator sets.

[0072] The specific expression for the energy storage lifetime depreciation cost is as follows:

[0073]

[0074] In the formula, This represents the total number of energy storage systems. The unit discharge capacity lifetime depreciation cost coefficient for energy storage; For energy storage during the period The discharge power.

[0075] The specific expression for interruptible load compensation cost is as follows:

[0076] In the formula, This represents the total number of interruptible loads. The unit power compensation price for interruptible loads; Interruptible load during time period Reduced power.

[0077] The specific expression for the interaction cost of the upper-level power grid is as follows:

[0078] In the formula, and Time periods Active power purchased and sold from the superior power grid; and Time periods The purchase price and the sale price of electricity.

[0079] Step 4: Establish an intraday phase optimization model with the objective function of minimizing the operational risk of the distribution network.

[0080]

[0081] In the formula, , , These are the risk indicators for node voltage exceeding limits, branch power exceeding limits, and load shedding. , , These are the weighting coefficients. System status The probability of; , , These are the total number of system nodes, the total number of system branches, and the total number of system load points, respectively. , , These represent the voltage amplitude at node i and the upper and lower limits of the voltage amplitude at node i, respectively. , These are the active power transmitted by line l and the maximum active power that line l is allowed to transmit, respectively. For load point Load demand; For load point The actual power supply.

[0082] Step 5: A novel global optimization algorithm is used to solve the two-stage optimization model.

[0083] like Figure 3 As shown, the novel global optimization algorithm includes the following specific steps:

[0084] Step 5.1: Within the search space, randomly generate a set of initial solutions based on the upper and lower bounds of the variables to form the search proxy set of the algorithm, so as to start the optimization process.

[0085]

[0086] In the formula, For the first The search agent in the 1st The position of the dimension; A random number that is uniformly distributed in the range [0,1]. and The first Upper and lower bounds of a dimensional variable.

[0087] Step 5.2: A residual-guided update mechanism is adopted, which aims to improve the accuracy of the initial search direction and accelerate convergence. The update process is as follows:

[0088] First, calculate the dynamic step size factor. :

[0089]

[0090] In the formula, where This represents the current iteration number. Total number of iterations

[0091] Then, calculate the gradient direction. :

[0092]

[0093] In the formula, A random number that is uniformly distributed in the range [0,1]. and Two bootstrap agents were selected based on residual probabilities; This is the current optimal position.

[0094] Finally, combining the Jacobian matrix Update proxy location Get a new position :

[0095]

[0096] Step 5.3: Implement a double-random reference update mechanism. This mechanism enhances the algorithm's global exploration capability by introducing randomness, thus avoiding getting trapped in local optima. Its position update formula is:

[0097] First, two reference proxy indexes are randomly selected. and obtain its location. , :

[0098]

[0099] Then, based on the random number Select update path:

[0100] when At that time, gradient updates are used to update the path and calculate the direction. And update location :

[0101]

[0102] when At that time, the Jacobian matrix is ​​used to update the path and calculate the direction. And update location :

[0103]

[0104] Step 5.4: Execute a weighted multi-path update mechanism. This mechanism balances the leaps and stability of the search by designing multiple update paths, enhancing local exploitation capabilities. In each dimension... Above, according to the conditions Dynamically select either path one or path two for execution. The position update formula is:

[0105] Path 1: Using random weights and Update location :

[0106]

[0107] Path 2: First, calculate the adaptive factor. Then use the intermediate results from step two. Update location :

[0108]

[0109] Step 5.5: Introduce the Lévy flight perturbation mechanism. Execute the Lévy flight perturbation based on the dynamically triggered probability to enable the algorithm to effectively escape local optima. The position update formula is as follows:

[0110] First, calculate the trigger probability that adapts to the iteration. If the conditions are met, then escape will be executed:

[0111]

[0112] Then, combine random numbers and elite agent positions Update proxy location :

[0113]

[0114] In the formula, For Levi's flight stride; and These are random numbers that are normally distributed. For the Lévy index parameters; For adaptive weights.

[0115] Step 5.6: Iteration and Output: Repeat steps 5.2 to 5.5 until the termination condition is met, and output the optimal scheduling scheme.

[0116] Step 6: Develop an optimized operation strategy for the distribution network based on the solution results. The optimal unit output, energy storage charging and discharging plan, and interruptible load dispatching scheme obtained from the optimization algorithm are then sent to the distribution network control system to achieve optimized control of the distribution network.

[0117] like Figure 4 The diagram shown is a structural schematic of a novel two-stage collaborative optimization device for power distribution networks that considers both economic efficiency and operational risks, according to another embodiment of the present invention. The device includes:

[0118] The parameter acquisition module is used to acquire distribution network parameters, distributed generation predicted output, and load forecast data.

[0119] The framework construction module is used to build the day-ahead and intraday two-stage optimization operation framework and constraints for the distribution network.

[0120] The day-ahead model building module is used to build a day-ahead stage optimization model, which takes minimizing the system operating cost as the objective function.

[0121] The intraday model building module is used to establish an intraday stage optimization model, which takes minimizing the operation risk of the distribution network as the objective function.

[0122] The optimization solution module is used to solve the two-stage optimization model using a novel global optimization algorithm;

[0123] The control and execution module is used to formulate optimal operation strategies for the distribution network based on the solution results and to control the distribution network.

[0124] Furthermore, the current-day model construction module is specifically used for:

[0125] A day-ahead optimization model is established with the objective function of minimizing system operating costs. The system operating costs include traditional unit operating costs, energy storage lifetime loss costs, interruptible load compensation costs, and interaction costs with the upstream power grid. The day-ahead optimization model is configured to satisfy general system operation constraints and system safety operation constraints, which include distribution network power flow balance constraints, unit output limit constraints, energy storage system operation constraints, interruptible load operation constraints, and system safety operation constraints.

[0126] Furthermore, the intraday model building module is specifically used for:

[0127] An intraday phase optimization model is established with the objective function of minimizing the operational risk of the distribution network. The operational risk is a weighted sum of node voltage over-limit risk, branch power over-limit risk, and load shedding risk. The intraday phase optimization model is configured to meet general system operation constraints, which include distribution network power flow balance constraints, generator output limit constraints, energy storage system operation constraints, and interruptible load operation constraints.

[0128] Furthermore, the optimization solution module is specifically used to execute the novel global optimization algorithm, including:

[0129] The algorithm comprises the following units: an initialization unit, which randomly generates an initial solution set within the search space based on the upper and lower bounds of variables, forming a search proxy set to initiate the optimization process; a guidance update unit, which executes a residual guidance update mechanism, utilizing elite individual information to construct a guidance vector, thereby improving the accuracy of the search direction and global exploration capability; a global exploration unit, which implements a double random reference update mechanism, introducing random reference individuals and gradient information during the exploration phase to dynamically balance the algorithm's exploration and development behaviors; a local development unit, which executes a weighted multi-path update mechanism, designing multi-path update strategies during the development phase to enhance the diversity and stability of local searches; a perturbation escape unit, which introduces a Lévy flight perturbation mechanism, executing Lévy flight perturbations based on dynamically triggered probabilities, enabling the algorithm to effectively escape local optima; and an iterative control unit, which determines whether the algorithm has reached the maximum number of iterations or convergence accuracy. If so, it outputs the optimal solution set; otherwise, it returns to the guidance update unit to continue iterative optimization.

[0130] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can realize the novel two-stage collaborative optimization method for power distribution networks that takes into account both economic efficiency and operational risks provided by any of the above-described method embodiments of the present invention.

[0131] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the device embodiment figures provided by this invention, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0132] Another preferred embodiment of the present invention provides a terminal device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a novel two-stage collaborative optimization method for power distribution networks that takes into account both economic efficiency and operational risks, as described in any of the above embodiments.

[0133] It should be noted that the terminal devices mentioned herein can be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. Terminal devices may include, but are not limited to, processors and memory. Those skilled in the art will understand that, for example, they may also include input / output devices, network access devices, buses, etc.

[0134] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting various parts of the terminal device via various interfaces and lines.

[0135] The memory can be used to store the computer program. The processor implements various functions of the terminal device by running or executing the computer program stored in the memory and calling data stored in the memory. The memory mainly includes a program storage area and a data storage area. The program storage area can store the operating system and at least one application program required for a function; the data storage area can store data created based on the use of the mobile phone. In addition, the memory can include high-speed random access memory and non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart memory card (SMC), secure digital card (SD card), flash memory card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0136] Another preferred embodiment of the present invention provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the storage medium is located to execute a novel two-stage collaborative optimization method for power distribution networks that takes into account both economic efficiency and operational risk, as described in any of the above embodiments.

[0137] The storage medium is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above-described methods. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0138] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A novel two-stage collaborative optimization method for power distribution networks that considers both economic efficiency and operational risks, characterized in that... include: (1) Obtain distribution network parameters, predicted output of distributed generation, and load forecast data; (2) Construct a two-stage optimized operation framework and system constraints for the distribution network from day-ahead to intraday; (3) Establish a day-ahead stage optimization model with the goal of minimizing system operating costs; (4) Establish an intraday phase optimization model with the goal of minimizing the operational risks of the distribution network; (5) A novel global optimization algorithm is used to solve the two-stage optimization model; (6) Develop a power distribution network optimization operation strategy based on the solution results.

2. The novel two-stage collaborative optimization method for power distribution networks considering both economic efficiency and operational risk, as described in claim 1, is characterized in that... The objective function of the current-day optimization model is: minC total =C gen +C ess +C il +C grid (1) In the formula, C gen For traditional unit operating costs; C ess Costs related to energy storage lifespan reduction; C il To compensate for interruptible load costs; C grid Costs associated with interaction with the upstream power grid.

3. The novel two-stage collaborative optimization method for power distribution networks considering both economic efficiency and operational risk, as described in claim 1, is characterized in that... The objective function of the intraday phase optimization model is: In the formula, R V (t), R P (t), R L (t) represents the node voltage over-limit risk, branch power over-limit risk, and load loss risk, respectively; γ1, γ2, and γ2 are weighting coefficients.

4. The novel two-stage collaborative optimization method for power distribution networks considering both economic efficiency and operational risk, as described in claim 1, is characterized in that... The novel global optimization algorithm includes the following steps: (1) Within the search space, a set of initial solutions is randomly generated based on the upper and lower bounds of the variables to form the search agent set of the algorithm, so as to start the optimization process; (2) A residual-guided update mechanism is adopted to construct a guiding vector using elite individual information, thereby improving the accuracy of the search direction and the global exploration capability; (3) Implement a double random reference update mechanism, introduce random reference individuals and gradient information in the exploration stage, and explore and develop dynamic balancing algorithms; (4) Implement a weighted multi-path update mechanism and design a multi-path update strategy during the development phase to enhance the diversity of local searches; (5) Introduce the Levy flight perturbation mechanism and execute the Levy flight perturbation according to the probability of dynamic triggering, so that the algorithm can effectively jump out of the local optimum region; (6) Determine whether the algorithm has reached the maximum number of iterations or convergence accuracy. If the conditions are met, output the optimal solution; otherwise, return to step (2) to continue iterative optimization.

5. A novel two-stage collaborative optimization method for power distribution networks considering both economic efficiency and operational risk, as described in claim 1, is characterized in that... The constraints of the optimization model include: The constraints of the day-ahead optimization model include: power flow balance constraints of the distribution network, power output limit constraints of generating units, operation constraints of the energy storage system, operation constraints of interruptible loads, and system safety operation constraints. The constraints of the intraday phase optimization model include: power flow balance constraints of the distribution network, power output limit constraints of generating units, operation constraints of the energy storage system, and operation constraints of interruptible loads.

6. A novel two-stage collaborative optimization device for power distribution networks that considers both economic efficiency and operational risks, characterized in that, include: The parameter acquisition module is used to acquire distribution network parameters, distributed generation predicted output, and load forecast data. The framework construction module is used to build the day-ahead and intraday two-stage optimization operation framework and constraints for the distribution network. The day-ahead model building module is used to build the day-ahead phase optimization model; The intraday model building module is used to build intraday phase optimization models. The optimization solution module is used to solve the model using a novel global optimization algorithm. The control and execution module is used to formulate optimal operation strategies for the distribution network based on the solution results and to control the distribution network.

7. A terminal device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium comprising a stored computer program, characterized in that, When the computer program is running, it controls the device containing the storage medium to implement the method as described in any one of claims 1 to 5.

Images