A CPU occupancy-based power meter data acquisition optimization method and system

Abstract

The present application relates to the field of smart meter data acquisition, especially to a CPU occupancy-based meter data acquisition optimization method and system, the method comprising: obtaining the signal strength and working state of concentrator and meter nodes, establishing a data acquisition network model and calculating the load rate to reflect the network busy degree. Vector coding is used to generate initial particle swarm, each particle represents a data acquisition path, the transmission efficiency and communication quality of the path are evaluated, and the path score is obtained by weighted fusion based on the load rate. The particle with the highest score is selected as the global optimal solution to guide the particle swarm learning, the learning ratio is adjusted through the negative exponential function, the parent node vector is updated and iteratively optimized until the score no longer improves, the optimal path is output, and the data acquisition path optimization is completed. The present application dynamically adjusts the weight and optimization strategy, adapts to network changes, reduces the hop count, avoids weak signal links, and improves the transmission efficiency and stability of the data acquisition path.

Description

Technical Field

[0001] This invention relates to the field of smart meter data acquisition. In particular, it relates to a method and system for optimizing smart meter data acquisition based on CPU utilization. Background Technology

[0002] In smart grid data acquisition systems, data collected from electricity meters needs to be transmitted from each meter node to the data terminal via communication links. The quality of the transmission path directly affects data acquisition efficiency and communication quality. Current data acquisition digital transmission optimization suffers from multi-objective conflicts, and traditional optimization algorithms have significant shortcomings in handling such problems: single-objective optimization algorithms (such as traditional particle swarm optimization) can only optimize a single objective (such as the shortest path) and cannot balance the multi-objective requirements of transmission efficiency and communication quality, resulting in problems such as communication instability or excessive transmission delay in practical applications.

[0003] Existing electricity meter data acquisition solutions mostly adopt "fixed paths" or "single index optimization". For example, if the path is selected based solely on signal strength, the following problems exist: When the electricity meter or data terminal has a high CPU utilization due to multitasking, the fixed path will cause data transmission delays, packet loss, and even affect the accuracy of basic metering; the network load (such as the number of active nodes) changes over time (such as peak electricity consumption and fault periods), and the fixed path cannot be adjusted in real time, resulting in a decline in the overall performance of the acquisition system. Summary of the Invention

[0004] To address the problems that existing multi-objective optimization algorithms use fixed weights, making it difficult to adapt to dynamic changes in network load, and that traditional particle swarm optimization algorithms are not well-suited to the tree-like topology of data acquisition networks, thus limiting the adaptability and effectiveness of path optimization, this invention provides solutions in the following aspects.

[0005] In the first aspect, a method for optimizing electricity meter data acquisition based on CPU occupancy includes: acquiring CPU occupancy data of data terminals and each electricity meter node, as well as the working status of each node; establishing a data acquisition network model and calculating the network load rate; representing paths in the data acquisition communication network using a vector encoding method, and generating an initial particle swarm based on the vector encoding, wherein each initial particle represents a data acquisition path; evaluating the parent node vector of the path represented by each particle in the initial particle swarm to determine its transmission efficiency and communication quality, and weighting and fusing the transmission efficiency and communication quality based on the network load rate to obtain the path score of each particle; selecting the particle with the highest path score as the global optimal solution, guiding the particle swarm to learn the particles of the global optimal solution, using a negative exponential function to perform exponential mapping on the ratio between the path score of each particle and the global optimal solution, and using this as the learning ratio of each particle, and updating the parent node vector; calculating the path score based on the updated parent node vector, updating the global optimal solution, iteratively updating the particle swarm until the path score of the global optimal solution no longer improves, terminating the iteration and outputting the parent node vector and path score corresponding to the current global optimal solution, thus completing the optimization of the data acquisition path.

[0006] By adopting the above technical solution, CPU utilization data and node working status are acquired in real time, and network load rate is dynamically calculated, which can accurately reflect the network busyness and adjust the path optimization strategy accordingly. Vector encoding is used to generate a valid initial particle swarm, and the transmission efficiency and communication quality of the path are comprehensively evaluated. The path score is obtained by weighted fusion based on network load rate to select the global optimal solution to guide the particle swarm learning. By dynamically adjusting the learning ratio, updating the parent node vector, and iterative optimization, this method can efficiently and stably complete the data acquisition path optimization, ensuring that the path achieves the best balance between transmission efficiency and communication quality, adapts to changes in network status, and provides a reliable optimization solution for the data acquisition system.

[0007] Preferably, the data acquisition network model includes a network node set comprising at least one data terminal node and multiple electricity meter nodes, with the data terminal as the root node and each electricity meter as a child node, and the link availability is evaluated based on a set signal strength threshold; wherein, the working state includes: active and idle.

[0008] Preferably, the network load rate is calculated using the following method:

[0009] The number of nodes in an active state is counted, the ratio of the number of active nodes to the total number of nodes is calculated, and then multiplied by a percentage to obtain the current network load rate of each node.

[0010] Preferably, the step of obtaining the data acquisition path includes:

[0011] The data acquisition network model is traversed using the BFS algorithm. Initial particles are generated based on the parent node vector encoding. Starting from the data terminal, a parent node is selected for each meter node that has not been assigned a parent node to ensure the legality of the path. This process continues until all nodes are assigned parent nodes, forming different initial parent node vectors. Each initial parent node vector corresponds to one particle, and each particle in the particle swarm corresponds to a data acquisition path in the communication network.

[0012] In this system, the index of each vector code corresponds to the node number, and the value of the vector code represents the parent node number of the corresponding node.

[0013] Preferably, the method for calculating the transmission efficiency includes:

[0014] The total number of hops is calculated by summing the hop counts of all nodes in the path of each particle from itself to the data terminal. The total number of hops is then exponentially decayed using a negative exponential function and normalized to obtain the transmission efficiency of each particle.

[0015] Preferably, the method for calculating the communication quality includes:

[0016] The difference between the weakest link signal strength in each particle's path and 1 is used as the signal offset. The signal offset is then divided by a preset signal strength threshold to obtain the communication quality score.

[0017] Preferably, the path score is calculated in the following ways:

[0018] The difference between the network load rate at the preset equilibrium point and the network load rate of the current particle is exponentially mapped. The result of the mapping is added to 1 and the reciprocal is taken as the weight of the transmission efficiency. The value of 1 minus the weight of the transmission efficiency is taken as the weight of the communication quality.

[0019] The path score for each particle is obtained by weighting the transmission efficiency and communication quality with their respective weights.

[0020] Preferably, the step of iteratively updating the particle swarm includes:

[0021] Obtain the learning ratio for each particle, and use the generated random number and the learning ratio to decide whether to copy the parent node vector elements of the global optimal solution or retain the elements of the current particle.

[0022] If the random number is less than the learning ratio, then copy the parent node vector elements of the global optimal solution; otherwise, if it is greater than or equal to the learning ratio, then retain the elements of the current particle and complete the particle update.

[0023] If the random number is less than the preset mutation probability, a mutation operation is performed. All elements that need to be mutated are randomly assigned a new parent node number and updated to the corresponding position in the parent node vector of the particle.

[0024] Perform a validity check on the path to ensure that the path is acyclic and conforms to the logic that the parent node number is less than the child node number. Otherwise, correct the parent node to the root node to ensure the correctness of the network topology and complete the particle update.

[0025] Secondly, a meter data acquisition optimization system based on CPU occupancy includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned meter data acquisition optimization method based on CPU occupancy is implemented.

[0026] The present invention has the following effects:

[0027] 1. This invention dynamically adjusts the weights of transmission efficiency and communication quality based on network load rate. It uses vector encoding to represent paths in the tree-like topology of the data acquisition network and combines it with a specific particle swarm optimization strategy. This allows the optimization strategy to be dynamically adjusted according to changes in network load rate, making data acquisition path optimization more flexible and adaptable to changes in the network environment. As a result, it can find better data acquisition paths under different network load conditions, thus improving the adaptability and robustness of data acquisition path optimization.

[0028] 2. This invention comprehensively considers two key factors: transmission efficiency and communication quality. It measures communication quality by the ratio of signal offset to a preset signal strength threshold and dynamically adjusts the weights of both factors based on network load. This allows the path score to more comprehensively and accurately reflect the quality of the path. By optimizing the data acquisition path, on the one hand, the number of path hops is reduced, thereby lowering data transmission latency and energy consumption, significantly improving transmission efficiency. On the other hand, path optimization avoids links with weak signal strength, effectively improving the reliability and accuracy of data transmission. This, in turn, enhances the overall performance of the data acquisition path, ensuring efficient and stable transmission of acquired data. Attached Figure Description

[0029] Figure 1 This is a flowchart of steps S1-S5 in an embodiment of the present invention, which describes an optimization method for electricity meter data acquisition based on CPU occupancy.

[0030] Figure 2 It is the set of network nodes in the data acquisition network model of an electricity meter data acquisition optimization method based on CPU utilization in an embodiment of the present invention.

[0031] Figure 3 This is a structural block diagram of an electricity meter data acquisition optimization system based on CPU utilization, according to an embodiment of the present invention. Detailed Implementation

[0032] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0033] Reference Figure 1 A method for optimizing electricity meter data acquisition based on CPU occupancy includes steps S1-S5, as detailed below:

[0034] S1: Obtain CPU occupancy data of data terminals and each meter node, as well as the working status of each node, establish a data acquisition network model, and calculate the network load rate to reflect the busy level of network data transmission.

[0035] The data acquisition module collects characteristic data from data terminals and all electricity meters within the data acquisition area. The data acquisition module includes a GPS (Global Positioning System) sensor, a communication module for the data terminal and electricity meters, and a log system.

[0036] By collecting the status of data terminals and electricity meters within the data acquisition area every hour, the system calculates real-time changes in network load rate, readjusts the target weights based on the real-time network load rate, calculates the path scores of all particles, and iteratively updates the global optimal solution until a new optimal data acquisition path is output. The data acquisition system can dynamically adjust the data acquisition path based on the real-time network load rate, ensuring a flexible balance between efficiency and quality. This adapts to dynamic changes in network status and improves the adaptability and stability of data acquisition path optimization.

[0037] The data acquisition network model includes a network node set consisting of at least one data terminal node and multiple electricity meter nodes, with the data terminal as the root node and each electricity meter as a child node, and the link availability is evaluated based on a set signal strength threshold; the working states include: active and idle.

[0038] It should be noted that a status monitoring module is deployed on the data terminal and each meter node to monitor the node's working status in real time. This status monitoring module can be a software or hardware module capable of detecting whether the node is transmitting data, waiting for acknowledgment, or whether the buffer queue is empty. The status monitoring module periodically (e.g., every second, every minute, or every hour) collects the node's working status and sends the status data to the data terminal. Each node's status data includes its node number and current working status (active or idle). Active: The node is transmitting data (sending / receiving frames), waiting for acknowledgment, or the buffer queue is not empty (data awaits transmission). Idle: The node has no data transmission needs and is in a sleep or listening state (only responding to wake-up commands).

[0039] The number of nodes in an active state is counted, that is, the number of nodes currently transmitting data, waiting for confirmation, or whose cache queue is not empty. The ratio of the number of active nodes to the total number of nodes is calculated and multiplied by a percentage to obtain the network load rate of each node. This is used to reflect the busy level of data transmission in the current network and to quantify the occupancy of node resources.

[0040] Network load factor influences the weighting of the objective function in data acquisition path optimization, especially when dynamically adjusting the weights of transmission efficiency and communication quality. By considering network load factor, the algorithm can more flexibly adapt to changes in network conditions, thereby finding the optimal data acquisition path under the current network load.

[0041] S2: Use vector encoding to represent paths in the data acquisition and communication network to ensure the effectiveness of network data transmission, and generate an initial particle swarm based on vector encoding, where each initial particle is a data acquisition path.

[0042] The data acquisition network model is traversed using the Breadth-First Search (BFS) algorithm. Initial particles are generated based on the parent node vector encoding. Starting from the data terminal, the parent node vector is used for encoding. A parent node is selected for each meter node that has not been assigned a parent node to ensure the legality of the path. This process continues until all nodes are assigned parent nodes, forming different initial parent node vectors. Each initial parent node vector corresponds to a particle, and each particle in the particle swarm corresponds to a data acquisition path in the communication network.

[0043] In this system, the index of each vector code corresponds to the node number, and the value of the vector code represents the parent node number of the corresponding node.

[0044] To ensure that the encoding corresponds to a valid data acquisition and transmission path, the parent node vector must meet the following constraints: for any meter node, its parent node must be effectively defined, traceable to the data terminal, and there must be no loops when tracing back along the parent node link, ultimately terminating at the data terminal (to prevent the data terminal from being unable to collect meter data).

[0045] In the generated data acquisition path, nodes are points in a tree topology, links are lines, and particles are the connection rules of the lines. This structure ensures that the data of each node can be transmitted to the data terminal through the node.

[0046] For example, the data terminal is set as the root node and the electricity meter is the child node. A wireless communication module is deployed on each child node to send and receive signals. Each node sends test signals periodically and records the received signal strength.

[0047] Reference Figure 2Define the set of network nodes in the data acquisition network model, for example, including one data terminal (root node) and 10 electricity meter nodes (child nodes). Use the preset signal strength as the characteristic parameter of the link and obtain the link with the lowest signal strength among all links. That is, a parent node (such as a data terminal or other electricity meter) may have communication links with multiple child nodes (such as an electricity meter), and each link between each child node and the parent node will have an independent signal strength value.

[0048] In the diagram, the parent node vector is encoded as Each number represents the parent node number of the corresponding node, clearly showing the hierarchical structure of the entire network. The root node, i.e., the concentrator E, has no parent node, so its parent node number is usually set to -1. The parent node numbers of other nodes correspond to the numbers of their directly connected upper-level nodes. For example, meter 1 (A) and meter 9 (F) are directly connected to the root node, so their parent node number is 0; meter 2 (B), meter 4 (C), and meter 7 (D) are connected to meter 1 (A), so their parent node number is 1, and so on. It should be noted that one tree corresponds to one parent node vector, and one parent node vector corresponds to one particle.

[0049] This invention requires that the parent node of any meter node be effectively defined to ensure that the path is traceable to the data terminal and that there are no loops when traversing back along the parent node link, ultimately terminating at the data terminal, thereby preventing data transmission failure. Furthermore, an initial particle swarm is generated based on parent node vector encoding. Starting from the data terminal, a BFS strategy is used to traverse the network, selecting a node that meets the signal strength threshold for each meter node without a assigned parent node, until all meter nodes are assigned parent nodes, forming the initial parent node vector.

[0050] S3: Evaluate the parent node vector of the path represented by each particle in the initial particle swarm to determine its transmission efficiency and communication quality. The transmission efficiency and communication quality are weighted and fused based on the network load rate to obtain the path score of each particle.

[0051] Methods for calculating transmission efficiency include:

[0052] The total number of hops is calculated by summing the hop counts of all nodes in the path of each particle from itself to the data terminal. The total number of hops is then exponentially decayed using a negative exponential function and normalized to obtain the transmission efficiency of each particle.

[0053] Calculate the sum of the path hops from each particle to the data terminal for all nodes in the path. The fewer the hops, the fewer intermediate nodes the data passes through, the lower the latency, the smaller the overhead, and the higher the efficiency. The transmission efficiency is inversely proportional to the total number of hops.

[0054] Methods for calculating communication quality include:

[0055] The difference between the weakest link signal strength in each particle's path and 1 is used as the signal offset. The signal offset is then divided by a preset signal strength threshold to obtain the communication quality score.

[0056] The higher the signal strength, the lower the link error rate and packet loss rate, and the more stable the communication. The overall communication quality of the path is determined by the signal strength of each link. Usually, the signal strength of the weakest link in the path is taken as the overall quality.

[0057] It is important to note that the preset signal strength threshold is a crucial parameter in wireless communication networks used to assess link availability. A link is considered usable only when its signal strength reaches or exceeds the preset level, thus ensuring communication quality. Furthermore, the signal strength threshold plays a vital role in network planning, optimization, and dynamic adjustment. Its setting directly impacts network performance and cost-effectiveness; thresholds that are too high or too low can adversely affect network stability and coverage. Therefore, setting a reasonable signal strength threshold is essential for maintaining the effective operation of wireless communication networks.

[0058] In this embodiment, the preset signal strength threshold is 85dB, which can be adjusted according to the specific environment and actual operating conditions of the network deployment.

[0059] The path score is calculated in the following ways:

[0060] The difference between the network load rate at the preset equilibrium point and the network load rate of the current particle is exponentially mapped. The result of the mapping is added to 1 and the reciprocal is taken as the weight of the transmission efficiency. The value of 1 minus the weight of the transmission efficiency is taken as the weight of the communication quality.

[0061] The path score for each particle is obtained by weighting the transmission efficiency and communication quality with their respective weights.

[0062] Specifically, the path score satisfies the following polynomial:

[0063] ;

[0064] ;

[0065] In the formula, Indicates the first The path score of each particle. Indicates the first The transport efficiency of individual particles This indicates the weight of transmission efficiency in the path score. Indicates the first The communication quality of individual particles, This indicates the weight of communication quality in the path score; The network load rate at the equilibrium point. This indicates the current network load rate of the particle.

[0066] When network load rate Approaching the equilibrium point hour, A value close to 0.5 indicates that transmission efficiency and communication quality are equally important. When the network load rate... When increasing, An increase indicates that transmission efficiency is becoming more important, because reducing the number of hops under high load can alleviate network congestion. When the network load rate... When decreasing, A decrease indicates that the importance of communication quality has increased, because signal strength has a more significant impact on transmission under low load.

[0067] By flexibly balancing transmission efficiency and communication quality based on real-time network load conditions, a score reflecting the quality of its path is calculated for each particle in the particle swarm, guiding the particle's search and optimization process in multidimensional space.

[0068] S4: Select the particle with the highest path score as the global optimal solution, guide the particle swarm to learn the particle of the global optimal solution, use a negative exponential function to perform exponential mapping on the ratio between the path score of each particle and the global optimal solution, and use it as the learning ratio of each particle, and update the parent node vector.

[0069] Specifically, the learning ratio satisfies the following relationship:

[0070] ;

[0071] in, Indicates the first The learning ratio of individual particles, Indicates the first The path score of each particle. This represents the globally optimal solution.

[0072] In other words, Reflects the first The path score difference of each particle is considered; a larger path score indicates a particle is closer to the global optimum, while a smaller score difference indicates a lower learning ratio. Furthermore, the learning ratio is considered within the range of... Within the range.

[0073] The update process includes the following steps:

[0074] Obtain the learning ratio for each particle, and use the generated random number and the learning ratio to decide whether to copy the parent node vector elements of the global optimal solution or retain the elements of the current particle.

[0075] Obtain the learning ratio for each particle, and use the generated random number and the learning ratio to decide whether to copy the parent node vector elements of the global optimal solution or retain the elements of the current particle.

[0076] If the random number is less than the preset mutation probability, a mutation operation is performed. All elements that need to be mutated are randomly assigned a new parent node number and updated to the corresponding position in the parent node vector of the particle.

[0077] Perform a validity check on the path to ensure that the path is acyclic and conforms to the logic that the parent node number is less than the child node number. Otherwise, correct the parent node to the root node to ensure the correctness of the network topology and complete the particle update.

[0078] If an invalid element exists, it will be forcibly corrected by changing the parent node of that element to the root node.

[0079] To further illustrate, for example, the preset mutation probability is 0.1, meaning that 10% of the nodes will be mutated. In order to enhance path diversity and prevent convergence to local optima, the number of each element in the parent node is randomly changed, which effectively balances global search and local exploitation and improves the algorithm's ability to find the optimal path.

[0080] S5: Based on the updated parent node vector, calculate the path score to update the global optimal solution, iteratively update the particle swarm until the path score of the global optimal solution no longer improves, terminate the iteration and output the parent node vector and path score corresponding to the current global optimal solution, thus completing the optimization of the data acquisition path.

[0081] This invention also provides a meter data acquisition optimization system based on CPU utilization. For example... Figure 3 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a method for optimizing electricity meter data acquisition based on CPU occupancy according to the first aspect of the present invention. The system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface. Their configuration and functions are known in the art and will not be described further here.

[0082] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for optimizing electricity meter data acquisition based on CPU utilization, characterized in that, include: Acquire CPU occupancy data of data terminals and each electricity meter node, as well as the working status of each node, establish a data acquisition network model, and calculate the network load rate; The paths in the data acquisition and communication network are represented using vector encoding, and an initial particle swarm is generated based on the vector encoding, where each initial particle is a data acquisition path. The parent node vector of the path represented by each particle in the initial particle swarm is evaluated to determine its transmission efficiency and communication quality. The transmission efficiency and communication quality are weighted and fused based on the network load rate to obtain the path score of each particle. The particle with the highest path score is selected as the global optimal solution. The particle swarm is then guided to learn from the global optimal solution. The ratio between the path score of each particle and the global optimal solution is used to perform an exponential mapping using a negative exponential function, which is then used as the learning ratio for each particle. The parent node vector is then updated. Based on the updated parent node vector, the path score is calculated to update the global optimum. The particle swarm is iteratively updated until the path score of the global optimum no longer improves. The iteration is terminated and the parent node vector and path score corresponding to the current global optimum are output, thus completing the optimization of the data acquisition path.

2. The method for optimizing electricity meter data acquisition based on CPU occupancy according to claim 1, characterized in that, The data acquisition network model includes a network node set consisting of at least one data terminal node and multiple electricity meter nodes, with the data terminal as the root node and each electricity meter as a child node, and the link availability is evaluated based on a set signal strength threshold; wherein, the working state includes: active and idle.

3. The method for optimizing electricity meter data acquisition based on CPU utilization according to claim 1, characterized in that, The network load rate is calculated in the following ways: The number of nodes in an active state is counted, the ratio of the number of active nodes to the total number of nodes is calculated, and then multiplied by a percentage to obtain the current network load rate of each node.

4. The method for optimizing electricity meter data acquisition based on CPU utilization according to claim 1, characterized in that, The steps for obtaining the data acquisition path include: The data acquisition network model is traversed using the BFS algorithm. Initial particles are generated based on the parent node vector encoding. Starting from the data terminal, a parent node is selected for each meter node that has not been assigned a parent node to ensure the legality of the path. This process continues until all nodes are assigned parent nodes, forming different initial parent node vectors. Each initial parent node vector corresponds to one particle, and each particle in the particle swarm corresponds to a data acquisition path in the communication network. In this system, the index of each vector code corresponds to the node number, and the value of the vector code represents the parent node number of the corresponding node.

5. The method for optimizing electricity meter data acquisition based on CPU occupancy according to claim 1, characterized in that, The calculation method for the transmission efficiency includes: The total number of hops is calculated by summing the hop counts of all nodes in the path of each particle from itself to the data terminal. The total number of hops is then exponentially decayed using a negative exponential function and normalized to obtain the transmission efficiency of each particle.

6. The method for optimizing electricity meter data acquisition based on CPU utilization according to claim 1, characterized in that, The methods for calculating the communication quality include: The difference between the weakest link signal strength in each particle's path and 1 is used as the signal offset. The signal offset is then divided by a preset signal strength threshold to obtain the communication quality score.

7. The method for optimizing electricity meter data acquisition based on CPU utilization according to claim 1, characterized in that, The path score is calculated in the following ways: The difference between the network load rate at the preset equilibrium point and the network load rate of the current particle is exponentially mapped. The result of the mapping is added to 1 and the reciprocal is taken as the weight of the transmission efficiency. The value of 1 minus the weight of the transmission efficiency is taken as the weight of the communication quality. The path score for each particle is obtained by weighting the transmission efficiency and communication quality with their respective weights.

8. The method for optimizing electricity meter data acquisition based on CPU utilization according to claim 1, characterized in that, The steps for iteratively updating the particle swarm include: Obtain the learning ratio for each particle, and use the generated random number and the learning ratio to decide whether to copy the parent node vector elements of the global optimal solution or retain the elements of the current particle. If the random number is less than the learning ratio, then copy the parent node vector elements of the global optimal solution; otherwise, if it is greater than or equal to the learning ratio, then retain the elements of the current particle and complete the particle update. If the random number is less than the preset mutation probability, a mutation operation is performed. All elements that need to be mutated are randomly assigned a new parent node number and updated to the corresponding position in the parent node vector of the particle. Perform a validity check on the path to ensure that the path is acyclic and conforms to the logic that the parent node number is less than the child node number. Otherwise, correct the parent node to the root node to ensure the correctness of the network topology and complete the particle update.

9. A meter data acquisition optimization system based on CPU utilization, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the electricity meter data acquisition optimization method based on CPU occupancy according to any one of claims 1-8.

Images