Equipment connection optimization method and system in substation centralized control system
Through the multi-ant colony optimization algorithm with a hybrid recommendation mechanism, the problems of slow convergence and local optimization in the optimization of device connection relationships are solved, the convergence speed and diversity of the algorithm are improved, and the flexibility of device connection relationships is realized. Efficient optimization.
Patent Information
- Application Number
- CN202411337201.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-25
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-09-25
AI Technical Summary
The ant colony algorithm has slow convergence, easy to fall into the problems of local optimization, limited adaptability and flexibility in the optimization of device connection relationships, which limits its application effect in complex and dynamic changing environments.
Multi-ant colony optimization algorithm using a hybrid recommendation mechanism includes content-based recommendation strategies, high-quality de-guided recommendation strategies based on convergence factors, browsing factor-based pheromone fusion recommendation strategies, and public path update recommendation strategies based on population similarity. Optimize the device connection relationship through self-organized mapping clustering algorithm and ACS algorithm, and dynamically adjust the pheromone update strategy to improve the convergence speed and diversity of the algorithm.
The performance of ant colony in complex optimization problems is improved, the algorithm converges to local optimal solutions prematurely, and the solution discovery ability in global optimization is enhanced, and the flexibility and efficient optimization of device connection relationships are achieved.
Smart Images

Figure CN118862942B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of electric power centralized control, and in particular to a method for optimizing the connection relationship between electric power equipment in a centralized control system of a substation. Background Art
[0002] With the continuous development of power systems and the improvement of intelligence levels, traditional power centralized control systems are facing new challenges and demands. In modern power monitoring systems, it is necessary to process massive amounts of data from various substations, including equipment operating conditions, environmental monitoring information, safety alarms, etc. In centralized control systems, the connection relationship between devices is one of the key factors to ensure the stable operation of the system. With the improvement of intelligence levels and the increase of system complexity, the connection relationship between devices has become more complex, which puts higher requirements on the management, maintenance and fault diagnosis of centralized control systems. The connection relationship between devices directly affects the data transmission efficiency. When the scale of centralized control systems expands and the number of devices increases, the connection relationship between devices and the amount of data generated by them increase sharply, which brings challenges to the management and optimization of device connection relationships. The device connection relationship in the centralized control system is not static and will change with factors such as the increase and decrease of devices and system upgrades. This requires that the management and optimization of device connection relationships have certain flexibility and adaptability. With the continuous evolution of network attack methods, ensuring the security of device connection relationships has become an important challenge. Effective security measures need to be taken to prevent data leakage and malicious attacks.
[0003] Reasonable connection relationships can reduce data transmission delays and improve the ability to process real-time data. Stable device connection relationships are the basis for ensuring the stable operation of the centralized control system. By optimizing the connection relationship, the system's anti-interference and fault tolerance capabilities can be improved, and the risk of system downtime caused by connection failures can be reduced. Therefore, when a fault occurs, a clear device connection relationship helps to quickly locate the fault point, simplify the fault diagnosis process, and shorten the recovery time.
[0004] Traditional equipment connection relationships in substations and industrial control systems usually rely on physical cables and static configuration methods, which leads to significant deficiencies in the system's scalability, flexibility, and maintenance costs. This method that relies on fixed physical connections faces high adjustment costs when equipment is added or the system is rearranged, while the static configuration management mode limits the system's ability to quickly adapt to changes. In addition, the centralized management architecture increases the vulnerability of the system. Once the central node encounters a problem, it may affect the operational stability of the entire system. The lack of effective redundant design and information island problems are also key factors that prevent traditional connection methods from meeting modern automation and intelligent needs. Therefore, exploring more flexible and efficient equipment connection solutions, such as introducing intelligent algorithms, has become an important requirement for improving system performance and reliability.
[0005] The ant colony algorithm provides a new idea for optimizing device connection relationships. However, in practical applications, its slow convergence, easy to fall into local optimality, limited adaptability and flexibility, etc., limit its application effect in certain complex and dynamically changing environments. Summary of the invention
[0006] The technical problem to be solved by the present invention is how to use the ant colony algorithm in the optimization application of device connection relationships, because its shortcomings such as slow convergence, easy to fall into local optimality, limited adaptability and flexibility limit its application effect in certain complex and dynamically changing environments.
[0007] The present invention solves the above technical problems through the following technical means:
[0008] The device connection optimization method in the substation centralized control system includes the following steps:
[0009] Step 1: Calculate the combined communication distance between devices ;
[0010] Step 2: Starting from the main device, a content-based recommendation strategy is used to generate well-distributed initial pheromones. First, a self-organizing map clustering algorithm is used to transform the edge device group into several small-scale edge sub-device groups. Based on the combined communication distance Use the ACS algorithm to find the shortest communication path from the master device to each edge sub-device group and give these communication paths pheromone rewards ;
[0011] Step 3: Enter the loop iteration phase and iterate the heterogeneous multi-ant colony ACS, MMAS and LRUACS in parallel;
[0012] Step 4: Calculate the browsing factor, update factor, and convergence factor on the iteration path;
[0013] Step 5: If the number of iterations is divisible by M, all populations communicate and execute step 6; otherwise, the trend of the iterative update result is analyzed to determine whether it is necessary to continue iterating. M is the iteration number threshold for determining the iteration effect.
[0014] Step 6: Implement a high-quality solution-guided recommendation strategy based on the convergence factor for ant colonies with slow convergence speed; when the browsing factor of an ant colony is large, it indicates that the diversity of the population is poor, and implement a pheromone fusion recommendation strategy based on the browsing factor for the ant colonies to improve the diversity of the population; implement a public communication path update recommendation strategy based on population similarity for the ant colonies to adjust the rewards and penalties of the public path;
[0015] Step 7: Adaptively update the pheromone of each path according to the relevant factors. If the iterative update stagnates, the circuit breaker mechanism is triggered to reset the pheromone of all paths, and the recommendation strategy based on association rules is used to adjust the pheromone reward of the communication path. And get the optimal communication path for each ant colony;
[0016] Step 8: When the maximum number of iterations is reached, the optimal communication path is output, and the connection relationship of the devices on the communication path is changed according to the optimal communication path from the main device to the target device.
[0017] Furthermore, in step 1, the combined communication distance The calculation process is as follows: First calculate the link system margin between devices i and j in the substation ,as well as
[0018]
[0019] In the formula is the signal transmission power, is the total link loss including free space loss, interface loss between devices, and antenna gain loss. is the minimum sensitivity of the receiving device;
[0020]
[0021] In the formula is the signal attenuation between devices i and j, is the signal-to-noise ratio between devices i and j, is the data transfer rate between devices i and j, , , , The combined communication distance , , , The weight ratio of A large constant to ensure that the combined communication distance is positive.
[0022] Furthermore, the sub-device group in step 2 is calculated as follows:
[0023]
[0024] Where n is the number of devices, is the number of sub-device groups, [] indicates rounding operation;
[0025] Use the ACS algorithm to find the shortest communication path for each sub-device group and give these communication paths pheromone rewards The calculation is as follows:
[0026]
[0027]
[0028] In the formula and are the optimal combined communication distance length and pheromone increment of the kth sub-device group, is the initial pheromone concentration of each communication path, It is the optimal communication path for a group of k sub-devices.
[0029] Furthermore, the step 3 comprises the following steps:
[0030] Step 3.1, each ant in device i of all ant colonies selects the next arrival device j;
[0031]
[0032] In the formula, is a random number in [0,1], The value range of is [0,1], is the pheromone concentration on the communication path between device i and device j, is the inverse of the combined distance between device i and device station j, is the expected heuristic factor, It is a collection of devices that have not arrived yet. For connected devices selected according to the roulette strategy;
[0033] Step 3.2: When each ant in the ACS ant colony and the LRUACS ant colony passes through the next device j from device i, it will send the local pheromone of the next iteration on the communication path. Make updates;
[0034]
[0035] In the formula is the evaporation rate of the local pheromone renewal strategy, is the number of iterations.
[0036] After all ants in the ACS ant colony complete the construction of the communication path, they update the global pheromone on the current global optimal communication path;
[0037]
[0038]
[0039] In the formula For pheromone increment, is the length of the current global optimal communication path, is the evaporation rate of the global pheromone update strategy, is the globally optimal communication path;
[0040] Step 3.3, MMAS updates the pheromone of the global optimal communication path or the shortest communication path generated in the current loop iteration;
[0041]
[0042]
[0043] in Indicates the current optimal solution or the global optimal solution;
[0044] Limit the pheromone concentration on all communication paths for the MMAS ant colony to prevent excessive accumulation of pheromones on certain communication paths, causing the algorithm to stagnate;
[0045]
[0046] is the maximum pheromone concentration on all communication paths, is the minimum pheromone concentration on all communication paths;
[0047] Step 3.4: The LRUACS ant colony performs global pheromone updates. First, the communication paths constructed by all ants are sorted, and the top 10% of the communication paths are selected. Then, the elite ants are updated with different degrees of pheromone according to the combined length of the communication paths. When the update factor of an edge reaches a certain threshold, the edge is no longer updated.
[0048]
[0049]
[0050] In the formula and Ant The length of the communication path and the pheromone increment, is the threshold of the cumulative update factor.
[0051] Furthermore, the step 4 comprises the following steps:
[0052] Step 4.1: In each iteration, record the number of times all ants pass through a communication path to form an n * n browsing matrix , find the browsing factor that reflects the distribution of communication paths chosen by ants to assess the diversity of each population ; When all ants only pass through a fixed communication path, the diversity of the algorithm is the worst, and the browsing factor is , the browsing factor is normalized to obtain ,The higher the value is, the more concentrated the communication paths sought by the ant colony is, indicating that the population diversity is worse.
[0053]
[0054]
[0055] Where n is the number of devices, It's an ant colony The browsing matrix;
[0056] Step 4.2: Calculate storage edges The update factor matrix of this iteration for the number of pheromone updates at and the cumulative update factor , and update the pheromone for the next iteration of the LRUACS ant colony;
[0057]
[0058]
[0059] Where t is the number of iterations, is the communication path of the pheromone to be updated, for The number of communication paths in
[0060] Step 4.3: Introduce the convergence factor representing the ratio of the global optimal solution to the average solution of each population , determine the guidance object and guidance intensity;
[0061]
[0062]
[0063] In the formula is the global optimal solution, For population No. The length of the optimal solution obtained by iteration, For population The convergence factor of is the convergence factor of the fastest converging population, is the difference in convergence factors;
[0064] Calculate the overall similarity of the optimal solution obtained by each group ;
[0065]
[0066] In the formula For population and The intersection of the optimal communication paths, i.e. the number of repeated edges, It is a population and The union of the optimal communication paths, i.e. the number of all edges;
[0067] Calculate population correlation coefficient .
[0068]
[0069] In the formula , and It stores the update factors of the optimal communication paths of the three populations: ACS, MMAS, and LRUACS.
[0070] Furthermore, the step 6 comprises the following steps:
[0071] Step 6.1: For ant colonies with slower convergence speed, a high-quality solution guidance recommendation strategy based on the convergence factor is implemented, and reward pheromones are used to gradually guide the communication path of high-quality solutions. The intensity of the reward is determined by the difference in the convergence factor.
[0072]
[0073] In the formula is a constant that controls the guiding strength.
[0074] Step 6.2: When the browsing factor of the ant colony is large, it indicates that the diversity of the population is poor. The pheromone fusion recommendation strategy based on the browsing factor is implemented for each ant colony to improve the diversity of the population.
[0075]
[0076] In the formula For population The pheromone matrix, To eliminate the population The pheromone matrix of other populations, H is the total number of populations, To eliminate the population Normalized browsing factor for other populations except
[0077] Step 6.3: Implement the recommended strategy for updating the public communication path based on population similarity for the ant colony to adjust the rewards and penalties of the public path. First, find the population and Number of public communication paths , and then the population and The contemporary update factor matrix and Added together, pheromones on common communication paths are finally dynamically rewarded or penalized based on population similarity.
[0078]
[0079]
[0080]
[0081] In the formula is a constant that controls the intensity of rewards or penalties.
[0082] The present invention also provides a device connection optimization system in a substation centralized control system, comprising:
[0083] Combined communication distance calculation module, used to calculate the combined communication distance between each device ;
[0084] The optimization module starts from the main device and uses a content-based recommendation strategy to generate well-distributed initial pheromones. First, the self-organizing map clustering algorithm is used to transform the edge device group into several small-scale edge sub-device groups. , use the ACS algorithm to find the shortest communication path for each sub-device group and give these communication paths pheromone rewards ;
[0085] Parallel iteration module, parallel iteration of heterogeneous multi-ant colony ACS, MMAS and LRUACS;
[0086] The multi-attribute decision-making module calculates the browsing factor, update factor, and convergence factor on the iteration path;
[0087] Iteration effect judgment module: if the number of iterations can be divided into M, all populations communicate and pass through the collaborative filtering recommendation module; otherwise, the trend of the iterative update results is analyzed to determine whether it is necessary to continue iterating. M is the iteration number threshold for judging the iteration effect;
[0088] The collaborative filtering recommendation module implements a high-quality solution-guided recommendation strategy based on the convergence factor for ant colonies with slow convergence speed; when the browsing factor of an ant colony is large, it indicates that the diversity of the population is poor, and a pheromone fusion recommendation strategy based on the browsing factor is implemented for the ant colonies to improve the diversity of the population; a public communication path update recommendation strategy based on population similarity is implemented for the ant colonies to adjust the rewards and penalties of the public path;
[0089] The communication path pheromone reward adjustment module adaptively updates the pheromone of each path according to the relevant factors. If the iterative update stagnates, the circuit breaker mechanism is triggered to reset the pheromone of all paths and use the recommendation strategy based on association rules to adjust the communication path pheromone reward. And get the optimal communication path for each ant colony;
[0090] The result output module outputs the optimal communication path when the maximum number of iterations is reached, and changes the connection relationship of the devices on the communication path according to the optimal communication path from the main device to the target device.
[0091] Furthermore, the optimization module sub-device population is calculated as follows:
[0092]
[0093] Where n is the number of devices, is the number of sub-device groups, [] indicates rounding operation;
[0094] Use the ACS algorithm to find the shortest communication path for each sub-device group and give these communication paths pheromone rewards The calculation is as follows:
[0095]
[0096]
[0097] In the formula and are the optimal combined communication distance length and pheromone increment of the kth sub-device group, is the initial pheromone concentration of each communication path, It is the optimal communication path for a group of k sub-devices.
[0098] Furthermore, the parallel iteration module execution process comprises the following steps:
[0099] Step 3.1, each ant in device i of all ant colonies selects the next arrival device j;
[0100]
[0101] In the formula, is a random number in [0,1], The value range of is [0,1], is the pheromone concentration on the communication path between device i and device j, is the inverse of the combined distance between device i and device station j, is the expected heuristic factor, It is a collection of devices that have not arrived yet. For connected devices selected according to the roulette strategy;
[0102] Step 3.2: When each ant in the ACS ant colony and the LRUACS ant colony passes through the next device j from device i, it will send the local pheromone of the next iteration on the communication path. Make updates;
[0103]
[0104] In the formula is the evaporation rate of the local pheromone renewal strategy, is the number of iterations.
[0105] After all ants in the ACS ant colony complete the construction of the communication path, they update the global pheromone on the current global optimal communication path;
[0106]
[0107]
[0108] In the formula For pheromone increment, is the length of the current global optimal communication path, is the evaporation rate of the global pheromone update strategy, is the globally optimal communication path;
[0109] Step 3.3, MMAS updates the pheromone of the global optimal communication path or the shortest communication path generated in the current loop iteration;
[0110]
[0111]
[0112] in Indicates the current optimal solution or the global optimal solution;
[0113] Limit the pheromone concentration on all communication paths for the MMAS ant colony to prevent excessive accumulation of pheromones on certain communication paths, causing the algorithm to stagnate;
[0114]
[0115] is the maximum pheromone concentration on all communication paths, is the minimum pheromone concentration on all communication paths;
[0116] Step 3.4: The LRUACS ant colony performs global pheromone updates. First, the communication paths constructed by all ants are sorted, and the top 10% of the communication paths are selected. Then, the elite ants are updated with different degrees of pheromone according to the combined length of the communication paths. When the update factor of an edge reaches a certain threshold, the edge is no longer updated.
[0117]
[0118]
[0119] In the formula and Ant The length of the communication path and the pheromone increment, is the threshold of the cumulative update factor.
[0120] Furthermore, the execution process of the multi-attribute decision-making module includes the following steps:
[0121] Step 4.1: In each iteration, record the number of times all ants pass through a communication path to form an n * n browsing matrix , find the browsing factor that reflects the distribution of communication paths chosen by ants to assess the diversity of each population ; When all ants only pass through a fixed communication path, the diversity of the algorithm is the worst, and the browsing factor is , the browsing factor is normalized to obtain ,The higher the value is, the more concentrated the communication paths sought by the ant colony is, indicating that the population diversity is worse.
[0122]
[0123]
[0124] Where n is the number of devices, It's an ant colony The browsing matrix;
[0125] Step 4.2: Calculate storage edges The update factor matrix of this iteration for the number of pheromone updates at and the cumulative update factor , and update the pheromone for the next iteration of the LRUACS ant colony;
[0126]
[0127]
[0128] Where t is the number of iterations, is the communication path of the pheromone to be updated, for The number of communication paths in
[0129] Step 4.3: Introduce the convergence factor representing the ratio of the global optimal solution to the average solution of each population , determine the guidance object and guidance intensity;
[0130]
[0131]
[0132] In the formula is the global optimal solution, For population No. The length of the optimal solution obtained by iteration, For population The convergence factor of is the convergence factor of the fastest converging population, is the difference in convergence factors;
[0133] Calculate the overall similarity of the optimal solution obtained by each group ;
[0134]
[0135] In the formula For population and The intersection of the optimal communication paths, i.e. the number of repeated edges, It is a population and The union of the optimal communication paths, i.e. the number of all edges;
[0136] Calculate population correlation coefficient .
[0137]
[0138] In the formula , and It stores the update factors of the optimal communication paths of the three populations: ACS, MMAS, and LRUACS.
[0139] The advantages of the present invention are:
[0140] 1. The present invention proposes a multi-ant colony optimization algorithm based on a hybrid recommendation mechanism, which has potential application prospects in the optimization of device connection relationships. By introducing a content-based recommendation strategy, three recommendation strategies are included: a high-quality solution guidance recommendation strategy based on a convergence factor: by identifying and guiding the ant colony to aggregate toward a known high-quality solution, the convergence of the algorithm is improved; a pheromone fusion recommendation strategy based on a browsing factor: by fusing pheromones from different sub-populations, the diversity of the search space is increased to avoid the algorithm from converging to a local optimal solution too early; a common path update recommendation strategy based on population similarity: by evaluating the similarity between different ant colonies and adaptively adjusting the pheromone update strategy, the exploration and utilization capabilities of the algorithm are balanced. This recommendation mechanism enables the algorithm to improve the convergence speed while maintaining diversity, thereby enhancing the performance of the ant colony in solving complex optimization problems.
[0141] 2. The present invention proposes a recommendation strategy based on association rules. When the algorithm falls into a local optimum and the search stagnates, this strategy allows the algorithm to adaptively update pheromones using the effective information accumulated in the previous search process, which helps the algorithm jump out of the local optimal solution and start exploring new potential optimization areas again. In this way, the algorithm can dynamically adjust the search strategy to avoid stagnation in a certain area for a long time, thereby finding a better solution in the global optimization problem. BRIEF DESCRIPTION OF THE DRAWINGS
[0142] Figure 1 The following is a logic diagram of the method execution in an embodiment of the present invention. DETAILED DESCRIPTION
[0143] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0144] This embodiment provides a method for optimizing the connection of equipment in a substation centralized control system. For example, the connection of equipment in a substation or a converter station can be optimized using the method of this embodiment. Figure 1 As shown, the method of this embodiment is performed in the following steps:
[0145] Step 1: Calculate the link system margin between devices i and j in the power system
[0146]
[0147] In the formula is the signal transmission power, is the total link loss including free space path loss, interface loss between devices, and antenna gain loss. is the minimum sensitivity of the receiving device.
[0148] In this embodiment, free space loss: This is the loss due to the attenuation of electromagnetic waves as the distance increases when propagating in free space. It can be calculated by the free space path loss formula, which is: SD = 20log10 (d) + 20log10 (f) - 147.55, where d is the distance (in meters) and f is the frequency (in Hertz). The 147.55 in the formula is a constant when using SI units of distance and frequency.
[0149] Interface loss between devices: This is the loss caused by the mismatch between device interfaces. This can usually be obtained by referring to the device specification or through actual measurement.
[0150] Antenna gain loss: Antenna gain refers to the multiple of the antenna's radiation intensity in a certain direction relative to an ideal omnidirectional radiating antenna. Antenna gain loss can be obtained by referring to the antenna specification or by actual measurement.
[0151] Minimum sensitivity of the receiving device: This refers to the minimum signal strength that the receiving device can receive. The minimum sensitivity of the receiving device can usually be obtained by referring to the device specification or through actual measurement.
[0152] Calculate the combined communication distance between devices In this embodiment, the combined communication distance is defined to measure the amount of resources consumed by the communication between two devices. The improved ant colony algorithm is subsequently used to obtain the optimal communication link between the main device and the edge device, thereby defining the device level and changing the device connection relationship.
[0153]
[0154] In the formula is the signal-to-noise ratio between devices i and j, is the data transfer rate between devices i and j, , , The combined communication distance , , The weight ratio of A large constant to ensure that the combined communication distance is positive.
[0155] In this embodiment, , , , use a spectrum analyzer to measure the received signal strength and noise intensity.
[0156] , is the channel bandwidth.
[0157] Since the link system margin is inversely proportional to the combined communication distance, a larger constant K is used to subtract the link system margin to make it proportional to the combined communication distance. At the same time, the weight ratio is defined based on the actual situation and the influence of each factor in the combined communication distance.
[0158] Step 2: Starting from the main device, a content-based recommendation strategy is used to generate well-distributed initial pheromones. First, the edge sub-device groups are divided according to the main device, that is, the device groups at the edge of the main device's communication radiation range. Clustering algorithms are used to divide individuals with similar properties in these groups into sub-groups and edge sub-device groups, ensuring that the number of devices in each edge sub-device group is within 10, so that ACS can find the optimal solution.
[0159]
[0160] Where n is the number of devices, is the number of edge sub-device groups, and [] indicates rounding operation.
[0161] Use the ACS algorithm to find the shortest communication path from the master device to each edge sub-device group and give these communication paths pheromone rewards .
[0162]
[0163]
[0164] In the formula and are the optimal combined communication distance length and pheromone increment of the kth sub-device group, is the initial pheromone concentration of each communication path, It is the optimal communication path for a group of k sub-devices.
[0165] Step 3: Enter the loop iteration phase and iterate the heterogeneous multi-ant colony ACS, MMAS and LRUACS in parallel.
[0166] Step 3.1: Each ant in device i of all ant colonies selects the next device to arrive at j.
[0167]
[0168] In the formula, is a random number in [0,1], The value range of is [0,1], is the pheromone concentration on the communication path between device i and device j, is the inverse of the combined distance between device i and device station j, is the expected heuristic factor, It is a collection of devices that have not arrived yet. For connected devices selected according to the roulette strategy.
[0169] Step 3.2: When each ant in the ACS ant colony and the LRUACS ant colony passes through the next device j from device i, it will send the local pheromone of the next iteration on the communication path. to update.
[0170]
[0171] In the formula is the evaporation rate of the local pheromone renewal strategy, is the number of iterations.
[0172] After all ants in the ACS ant colony complete the construction of the communication path, they update the global pheromone on the current global optimal communication path.
[0173]
[0174]
[0175] In the formula For pheromone increment, is the length of the current global optimal communication path, is the evaporation rate of the global pheromone update strategy, is the globally optimal communication path.
[0176] Step 3.3: MMAS updates the pheromone of the global optimal communication path or the shortest communication path generated in the current loop iteration.
[0177]
[0178]
[0179] in Indicates the current optimal solution or the global optimal solution.
[0180] The pheromone concentration on all communication paths of the MMAS ant colony is limited to prevent excessive accumulation of pheromones on certain communication paths and cause the algorithm to stagnate.
[0181]
[0182] is the maximum pheromone concentration on all communication paths, is the minimum pheromone concentration on all communication paths.
[0183] Step 3.4: The LRUACS ant colony performs global pheromone updates. First, the communication paths constructed by all ants are sorted, and the top 10% of the communication paths are selected. Then, the elite ants are updated with different degrees of pheromone according to the combined length of the communication paths. When the update factor of an edge reaches a certain threshold, the edge is no longer updated.
[0184]
[0185]
[0186] In the formula and Ant The length of the communication path and the pheromone increment, is the threshold of the cumulative update factor.
[0187] Step 4: Build a multi-attribute decision-making model and calculate the browsing factor, update factor, and convergence factor on the iterative path.
[0188] Step 4.1: In each iteration, record the number of times all ants pass through a communication path to form an n * n browsing matrix , find the browsing factor that reflects the distribution of communication paths chosen by ants to assess the diversity of each population When all ants only pass through a fixed communication path, the diversity of the algorithm is the worst, and the browsing factor is , the browsing factor is normalized to obtain ,The higher the value is, the more concentrated the communication paths sought by the ant colony is, indicating that the population diversity is worse.
[0189]
[0190]
[0191] Where n is the number of devices, It's an ant colony The browse matrix.
[0192] Step 4.2: Calculate storage edges The update factor matrix of this iteration for the number of pheromone updates at and the cumulative update factor , and update the pheromone for the next iteration of the LRUACS ant colony.
[0193]
[0194]
[0195] Where t is the number of iterations, is the communication path of the pheromone to be updated, for The number of communication paths in the .
[0196] Step 4.3: Introduce the convergence factor representing the ratio of the global optimal solution to the average solution of each population , determine the guidance object and guidance intensity.
[0197]
[0198]
[0199] In the formula is the global optimal solution, For population No. The length of the optimal solution obtained by iteration, For population The convergence factor of is the convergence factor of the fastest converging population, is the difference in convergence factors. When the difference in convergence factors is greater than 1%, it means that the population converges slowly and requires a high-quality solution to guide it.
[0200] Calculate the overall similarity of the optimal solution obtained by each group .
[0201]
[0202] In the formula For population and The intersection of the optimal communication paths, i.e. the number of repeated edges, It is a population and The union of the optimal communication paths, that is, the number of all edges.
[0203] Calculate population correlation coefficient .
[0204]
[0205] In the formula , and It stores the update factors of the optimal communication paths of the three populations: ACS, MMAS, and LRUACS.
[0206] Step 5: If the number of iterations is divisible by M, all populations communicate and pass through the collaborative filtering recommendation module. Otherwise, the trend of the iterative update results is analyzed to determine whether it is necessary to continue iterating. M is the iteration number threshold for judging the iteration effect.
[0207] Step 6: Build a collaborative filtering recommendation module.
[0208] Step 6.1: For ant colonies with slower convergence speed, a high-quality solution guidance recommendation strategy based on the convergence factor is implemented, and reward pheromones are used to gradually guide the communication path of high-quality solutions. The intensity of the reward is determined by the difference in the convergence factor.
[0209]
[0210] In the formula is a constant that controls the guiding strength.
[0211] Step 6.2: When the browsing factor of the ant colony is large, it indicates that the diversity of the population is poor. The pheromone fusion recommendation strategy based on the browsing factor is implemented for each ant colony to improve the diversity of the population.
[0212]
[0213] In the formula For population The pheromone matrix, To eliminate the population The pheromone matrix of other populations, H is the total number of populations, To eliminate the population Normalized browsing factor for other populations except
[0214] Step 6.3: Implement the recommended strategy for updating the public communication path based on population similarity for the ant colony to adjust the rewards and penalties of the public path. First, find the population and Number of public communication paths , and then the population and The contemporary update factor matrix and Added together, pheromones on common communication paths are finally dynamically rewarded or penalized based on population similarity.
[0215]
[0216]
[0217]
[0218] In the formula is a constant that controls the intensity of rewards or penalties.
[0219] Step 7: Adaptively update the pheromone of each path according to the relevant factors. If the iterative update stagnates, the circuit breaker mechanism is triggered to reset the pheromone of all paths, and the recommendation strategy based on association rules is used to adjust the pheromone reward of the communication path. And get the optimal communication path for each ant colony.
[0220]
[0221]
[0222] In the formula is a constant that controls the update intensity, is the adaptive coefficient.
[0223] Step 8: When the maximum number of iterations is reached, the optimal communication path is output, and the connection relationship of the devices on the communication path is changed according to the optimal communication path from the main device to the target device.
[0224] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for optimizing equipment connection in a substation centralized control system, characterized in that: The following steps are involved: Step 1: Calculate the combined communication distance between devices ; Step 2: Starting from the main device, a content-based recommendation strategy is used to generate well-distributed initial pheromones. First, a self-organizing map clustering algorithm is used to convert the edge device group into several small-scale edge sub-device groups. Based on the combined communication distance , use the ACS algorithm to find the shortest communication path from the master device to each edge sub-device group and give these communication paths pheromone rewards ; Step 3: Enter the loop iteration phase and iterate the heterogeneous multi-ant colony ACS, MMAS and LRUACS in parallel; Step 4: Calculate the browsing factor, update factor, and convergence factor on the iterative path; Step 5: If the number of iterations is divisible by M, all populations communicate and execute step 6. Otherwise, the trend of the iterative update result is analyzed to determine whether further iteration is needed. M is the iteration number threshold for judging the iteration effect. Step 6: For ant colonies with slow convergence speed, a high-quality solution-guided recommendation strategy based on the convergence factor is implemented. When the browsing factor of an ant colony is large, indicating that the diversity of the population is poor, a pheromone fusion recommendation strategy based on the browsing factor is implemented for each ant colony to improve the diversity of the population. For ant colonies, a public communication path update recommendation strategy based on population similarity is implemented to adjust the rewards and penalties of the public path. Step 7: Adaptively update the pheromone of each path according to the relevant factors. If the iterative update stagnates, the circuit breaker mechanism is triggered to reset the pheromone of all paths and use the recommendation strategy based on association rules to adjust the communication path pheromone reward. And get the optimal communication path for each ant colony; Step 8: When the maximum number of iterations is reached, the optimal communication path is output, and the connection relationship of the devices on the communication path is changed according to the optimal communication path from the master device to the target device.
2. The device connection optimization method in the substation centralized control system according to claim 1 is characterized in that: In step 1, the combined communication distance The calculation process is as follows: First calculate the link system margin between devices i and j in the substation ,as well as In the formula is the signal transmission power, is the total link loss including free space path loss, interface loss between devices, and antenna gain loss. is the minimum sensitivity of the receiving device; In the formula is the signal-to-noise ratio between devices i and j, is the data transmission rate between devices i and j, 、 、 The combined communication distance 、 、 The weight ratio of A large constant to ensure that the combined communication distance is positive.
3. The device connection optimization method in the substation centralized control system according to claim 2 is characterized in that: The sub-device group in step 2 is calculated as follows: Where n is the number of devices, is the number of sub-device groups, [] indicates rounding operation; Use the ACS algorithm to find the shortest communication path for each sub-device group and give these communication paths pheromone rewards The calculation is as follows: In the formula and are the optimal combined communication distance length and pheromone increment of the kth sub-device group, is the initial pheromone concentration of each communication path, It is the optimal communication path for a group of k sub-devices.
4. The device connection optimization method in the substation centralized control system according to claim 3 is characterized in that: The step 3 comprises the following steps: Step 3.1: Each ant in device i of all ant colonies selects the next arriving device j; Where, is a random number in [0,1], The value range of is [0,1], is the pheromone concentration on the communication path between device i and device j, is the inverse of the combined distance between device i and device station j, is the expected heuristic factor, It is a collection of devices that have not arrived yet. For connected devices selected according to roulette strategy; Step 3.2: When each ant in the ACS ant colony and the LRUACS ant colony passes through the next device j from device i, it will send the local pheromone of the next iteration on the communication path. Make updates; In the formula is the evaporation rate of the local pheromone renewal strategy, is the number of iterations; After all ants in the ACS ant colony complete the construction of the communication path, they update the global pheromone on the current global optimal communication path; In the formula is the pheromone increment, is the length of the current global optimal communication path, is the evaporation rate of the global pheromone update strategy, is the globally optimal communication path; Step 3.3, MMAS updates the pheromone of the global optimal communication path or the shortest communication path generated in the current loop iteration; in Indicates the current optimal solution or the global optimal solution; Limit the pheromone concentration on all communication paths of the MMAS ant colony to prevent excessive accumulation of pheromones on certain communication paths, which may cause the algorithm to stagnate. is the maximum pheromone concentration on all communication paths, is the minimum pheromone concentration on all communication paths; Step 3.4: The LRUACS ant colony performs a global pheromone update. First, the communication paths constructed by all ants are sorted, and the top 10% of communication paths are selected. Then, the elite ants are updated with different degrees of pheromone based on the combined length of the communication paths. When the update factor of an edge reaches a certain threshold, the edge is no longer updated. In the formula and Ants The length of the communication path and the pheromone increment, is the threshold of the cumulative update factor.
5. The device connection optimization method in the substation centralized control system according to claim 4 is characterized in that: The step 4 comprises the following steps: Step 4.1: In each iteration, record the number of times all ants pass through a communication path to form an n * n browsing matrix , find the browsing factor that reflects the distribution of communication paths chosen by ants to evaluate the diversity of each population ; When all ants only pass through a fixed communication path, the diversity of the algorithm is the worst, and the browsing factor is , the browsing factor is normalized to obtain ,The higher the value is, the more concentrated the communication paths sought by the ant colony, indicating a poorer population diversity; Where n is the number of devices, It's an ant colony The browsing matrix; Step 4.2: Calculate storage edges The update factor matrix of this iteration for the number of pheromone updates at and cumulative update factor , and update the pheromone for the next iteration of the LRUACS ant colony; Where t is the number of iterations, is the communication path of the pheromone to be updated, for The number of communication paths in Step 4.3: Introduce the convergence factor representing the ratio of the global optimal solution to the average solution of each population , determine the guidance object and guidance intensity; In the formula is the global optimal solution, For population No. The length of the optimal solution obtained by iteration, For population The convergence factor, is the convergence factor of the fastest-converging population, is the difference in convergence factors; Calculate the overall similarity of the optimal solution obtained by each group ; In the formula For population and The intersection of the optimal communication paths, i.e. the number of repeated edges, It is a population and The union of the optimal communication paths, i.e. the number of all edges; Calculate population correlation coefficient ; In the formula 、 and It stores the update factors of the optimal communication paths of the three populations ACS, MMAS, and LRUACS.
6. The device connection optimization method in the substation centralized control system according to claim 5 is characterized in that: The step 6 comprises the following steps: Step 6.1: Implement a high-quality solution recommendation strategy based on the convergence factor for ant colonies with slower convergence speeds. Use reward pheromones to gradually guide them along the communication path of high-quality solutions. The intensity of the reward is determined by the difference in the convergence factor. In the formula is a constant that controls the guiding strength; Step 6.2: When the browsing factor of an ant colony is large, indicating that the diversity of the population is poor, a pheromone fusion recommendation strategy based on the browsing factor is implemented for each ant colony to improve the diversity of the population; In the formula For population The pheromone matrix, To eliminate the population The pheromone matrix of other populations, H is the total number of populations, To eliminate the population Normalized browsing factors of other populations except Step 6.3: Implement the public communication path update recommendation strategy based on population similarity for the ant colony to adjust the rewards and penalties of the public path. First, find the population and Number of public communication paths , and then the population and The contemporary update factor matrix and Add them together and finally dynamically reward or penalize pheromones on the common communication path based on population similarity; In the formula is a constant that controls the intensity of rewards or penalties.
7. The equipment connection optimization system in the substation centralized control system is characterized by: include: Combined communication distance calculation module, used to calculate the combined communication distance between devices ; The optimization module starts from the main device and uses a content-based recommendation strategy to generate well-distributed initial pheromones. First, the self-organizing map clustering algorithm is used to transform the edge device group into several small-scale edge sub-device groups. Based on the combined communication distance , use the ACS algorithm to find the shortest communication path for each sub-device group and give these communication paths pheromone rewards ; Parallel iteration module, parallel iteration of heterogeneous multi-ant colony ACS, MMAS and LRUACS; Multi-attribute decision-making module calculates the browsing factor, update factor, and convergence factor on the iterative path; Iteration effect judgment module: if the number of iterations is divisible by M, all populations communicate and pass through the collaborative filtering recommendation module. Otherwise, the trend of the iterative update results is analyzed to determine whether to continue iteration. M is the iteration number threshold for judging the iteration effect. The collaborative filtering recommendation module implements a high-quality solution-guided recommendation strategy based on the convergence factor for ant colonies with slow convergence speeds. When the browsing factor of an ant colony is large, indicating that the diversity of the population is poor, a pheromone fusion recommendation strategy based on the browsing factor is implemented for each ant colony to improve the diversity of the population. A public communication path update recommendation strategy based on population similarity is implemented for ant colonies to adjust the rewards and penalties of public paths. The communication path pheromone reward adjustment module adaptively updates the pheromone of each path according to the relevant factors. If the iterative update stagnates, the circuit breaker mechanism is triggered to reset the pheromone of all paths and use the recommendation strategy based on association rules to adjust the communication path pheromone reward. And get the optimal communication path for each ant colony; The result output module outputs the optimal communication path when the maximum number of iterations is reached, and changes the connection relationship of the devices on the communication path according to the optimal communication path from the main device to the target device.
8. The device connection optimization system in the substation centralized control system according to claim 7 is characterized in that: The optimization module sub-device population is calculated as follows: Where n is the number of devices, is the number of sub-device groups, [] indicates rounding operation; Use the ACS algorithm to find the shortest communication path for each sub-device group and give these communication paths pheromone rewards The calculation is as follows: In the formula and are the optimal combined communication distance length and pheromone increment of the kth sub-device group, is the initial pheromone concentration of each communication path, It is the optimal communication path for a group of k sub-devices.
9. The device connection optimization system in the substation centralized control system according to claim 8, characterized in that: The parallel iteration module execution process includes the following steps: Step 3.1: Each ant in device i of all ant colonies selects the next arriving device j; Where, is a random number in [0,1], The value range of is [0,1], is the pheromone concentration on the communication path between device i and device j, is the inverse of the combined distance between device i and device station j, is the expected heuristic factor, It is a collection of devices that have not arrived yet. For connected devices selected according to roulette strategy; Step 3.2: When each ant in the ACS ant colony and the LRUACS ant colony passes through the next device j from device i, it will send the local pheromone of the next iteration on the communication path. Make updates; In the formula is the evaporation rate of the local pheromone renewal strategy, is the number of iterations; After all ants in the ACS ant colony complete the construction of the communication path, they update the global pheromone on the current global optimal communication path; In the formula is the pheromone increment, is the length of the current global optimal communication path, is the evaporation rate of the global pheromone update strategy, is the globally optimal communication path; Step 3.3, MMAS updates the pheromone of the global optimal communication path or the shortest communication path generated in the current loop iteration; in Indicates the current optimal solution or the global optimal solution; Limit the pheromone concentration on all communication paths of the MMAS ant colony to prevent excessive accumulation of pheromones on certain communication paths, which may cause the algorithm to stagnate. is the maximum pheromone concentration on all communication paths, is the minimum pheromone concentration on all communication paths; Step 3.4: The LRUACS ant colony performs a global pheromone update. First, the communication paths constructed by all ants are sorted, and the top 10% of communication paths are selected. Then, the elite ants are updated with different degrees of pheromone based on the combined length of the communication paths. When the update factor of an edge reaches a certain threshold, the edge is no longer updated. In the formula and Ants The length of the communication path and the pheromone increment, is the threshold of the cumulative update factor.
10. The device connection optimization system in the substation centralized control system according to claim 9, characterized in that: The execution process of the multi-attribute decision-making module includes the following steps: Step 4.1: In each iteration, record the number of times all ants pass through a communication path to form an n * n browsing matrix , find the browsing factor that reflects the distribution of communication paths chosen by ants to evaluate the diversity of each population ; When all ants only pass through a fixed communication path, the diversity of the algorithm is the worst, and the browsing factor is , the browsing factor is normalized to obtain ,The higher the value is, the more concentrated the communication paths sought by the ant colony, indicating a poorer population diversity; Where n is the number of devices, It's an ant colony The browsing matrix; Step 4.2: Calculate storage edges The update factor matrix of this iteration for the number of pheromone updates at and cumulative update factor , and update the pheromone for the next iteration of the LRUACS ant colony; Where t is the number of iterations, is the communication path of the pheromone to be updated, for The number of communication paths in Step 4.3: Introduce the convergence factor representing the ratio of the global optimal solution to the average solution of each population , determine the guidance object and guidance intensity; In the formula is the global optimal solution, For population No. The length of the optimal solution obtained by iteration, For population The convergence factor, is the convergence factor of the fastest-converging population, is the difference in convergence factors; Calculate the overall similarity of the optimal solution obtained by each group ; In the formula For population and The intersection of the optimal communication paths, i.e. the number of repeated edges, It is a population and The union of the optimal communication paths, i.e. the number of all edges; Calculate population correlation coefficient ; In the formula 、 and It stores the update factors of the optimal communication paths of the three populations ACS, MMAS, and LRUACS.
