Multi-scale coordination control method

Abstract

The invention relates to a multi-scale coordination control method, which belongs to the field of power system control and comprises the following steps: firstly, acquiring multi-scale data of a controlled system; performing scale feature analysis on the data to identify each scale behavior pattern and association relationship; based on the analysis result, generating a multi-scale coordination control strategy which comprehensively considers the interaction between the scales; a control instruction is distributed and executed through a coordinator, and dynamic adjustment is carried out to ensure that the multi-scale targets are consistent; and finally monitoring the control effect in real time and feeding back to form closed-loop self-adaptive control. According to the method, by coordinating control actions of different scales such as macroscopic scale, mesoscopic scale and microscopic scale, the scale target conflict problem existing in a traditional method is effectively solved, and the overall performance, the robustness and the self-adaptive capacity of the system are remarkably improved.

Description

Technical Field

[0001] This invention relates to a multi-scale coordinated control method, belonging to the field of power system control. Background Technology

[0002] With the development of technology, the complexity of modern engineering systems (such as smart grids, smart manufacturing, and large-scale Internet of Things) is increasing day by day. These systems typically exhibit significant multi-scale characteristics, that is, on the time scale, there are both rapid dynamic processes on the order of seconds or even milliseconds and slow evolutionary processes on the order of hours or days; on the spatial scale, they cover the micro level from the individual device / sensor, to the meso level of the workshop / regional power grid, and then to the macro level of the entire factory / wide area power grid.

[0003] Currently, most traditional control methods are designed for single-scale or isolated subsystems. For example, some controllers focus on the rapid response of local devices, while others are responsible for steady-state optimization across the entire system. This "divide and conquer" strategy has revealed serious limitations in practical applications: First, control objectives at different scales (such as local optima and global optima) often conflict, leading to a decline in overall system performance and creating a situation where each system operates independently. Second, due to the lack of effective cross-scale information exchange and coordination mechanisms, control actions at one scale may cause unpredictable negative interference to another scale, or even trigger cascading failures. Third, when the system is subjected to internal or external disturbances, traditional methods struggle to achieve coordinated recovery of multi-scale states, resulting in insufficient system robustness and adaptability.

[0004] Therefore, there is an urgent need in this field for an integrated method that can fundamentally break down scale barriers and achieve multi-level, multi-time-dimensional collaborative decision-making and control at the macro, meso, and micro levels. This method aims to coordinate control behaviors at different scales from a global system perspective, thereby optimizing the overall system performance while ensuring local performance. Summary of the Invention

[0005] Based on the problems described in the background, the problem to be solved by the present invention is to provide a multi-scale coordinated control method to address the problems mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a multi-scale coordinated control method, comprising the following steps: (1) Obtain multi-scale data from the controlled system, wherein the multi-scale data includes system state information at at least two different scales, wherein the scale includes at least one of time scale and spatial scale, and each scale corresponds to a different data acquisition frequency or resolution; (2) Perform scale feature analysis on the multi-scale data to identify system behavior patterns and relationships at each scale, wherein the scale feature analysis includes data fusion, pattern extraction and anomaly detection; (3) Based on the results of the scale feature analysis, a multi-scale coordinated control strategy is generated. The control strategy includes a set of control instructions for different scales, wherein the interaction and priority between scales are considered when generating the control strategy. (4) Execute the multi-scale coordinated control strategy, distribute control commands to the corresponding execution units of the controlled system through the coordinator, and dynamically adjust the control actions during the execution process to ensure the consistency of control objectives at different scales; (5) Monitor the control effect in real time and feed back the monitoring results to steps (1) to (4) to achieve closed-loop adaptive control.

[0007] Preferably, the multi-scale data in step (1) includes macro-scale data, meso-scale data and micro-scale data, wherein macro-scale data reflects the overall operating status of the system, meso-scale data reflects the status of subsystems or regions, and micro-scale data reflects the status of local components or real-time details; and the data acquisition is achieved through at least one of sensor networks, historical databases or simulation.

[0008] Preferably, the scale feature analysis in step (2) includes: Normalize multi-scale data to eliminate the impact of scale differences; Employ machine learning algorithms or statistical analysis methods to uncover potential patterns and trends in data at various scales; The strength of coupling between scales is evaluated, and a scale correlation matrix is ​​generated for subsequent control strategy generation. The scale feature analysis can identify the stability and vulnerability indicators of the system at multiple scales.

[0009] Preferably, when generating the multi-scale coordinated control strategy in step (3), a hierarchical decision-making mechanism is adopted, including: Set long-term or overall control objectives at a macro scale; Develop medium-term or regional control plans at the meso-scale; Formulate short-term or localized control actions at the microscale; The control strategy is generated through an optimization algorithm to minimize inter-scale conflicts and maximize system efficiency, and the control commands include setpoint adjustment, mode switching, or parameter tuning.

[0010] Preferably, in step (4), the coordinator uses an event-driven or time-triggered mechanism to distribute control commands, and the coordination process includes: Control resources are dynamically allocated based on scale priority. Detect and resolve control conflicts between scales, and ensure the coordination of control actions through arbitration logic; Delay compensation and fault tolerance are introduced into control execution to deal with communication or execution anomalies; The coordinator implements intelligent decision-making based on fuzzy logic or a rule engine.

[0011] Preferably, the multi-scale data includes grid frequency, voltage level, load demand and renewable energy output, and the control objectives include frequency stability, voltage optimization and energy dispatch; the monitoring feedback in step (5) is linked with the grid dispatch center to achieve multi-timescale power generation and power consumption coordination.

[0012] Preferably, when generating the control strategy in step (3), a predictive control step is also included: Based on historical data and real-time information, predict the state evolution of the system at multiple scales in the future; Adjust control strategies in advance based on forecast results to cope with potential disturbances or changes in demand; The predictive control adopts a rolling optimization method and is combined with the feedback mechanism in step (5) to improve control accuracy.

[0013] Preferably, the real-time monitoring in step (5) includes performance index calculation, which includes response time, energy consumption, stability or efficiency; and the feedback adjustment adopts an adaptive learning algorithm to automatically update the control parameters according to system changes. The adaptive learning algorithm includes neural networks, genetic algorithms or reinforcement learning.

[0014] Preferably, it also includes a scale priority management step: Priority weights are assigned to different scales, with the priorities dynamically adjusted based on system criticality, real-time requirements, or external events. In the generation and execution of control strategies, high-scale control tasks are given priority. The priority management is implemented through weighted voting or consensus algorithms to ensure that the core functions of the system are not affected when resources are limited.

[0015] Preferably, the method is applied to an intelligent manufacturing system, wherein the multi-scale data includes equipment status, production progress, material flow and energy consumption, and the control objectives include maximizing production efficiency, fault prevention and resource optimization; the coordinated execution in step (4) is integrated with the manufacturing execution system to realize multi-scale collaborative control from the factory level to the equipment level.

[0016] The beneficial effects of this invention are: 1. By performing scale feature analysis and generating multi-scale coordinated control strategies, this invention can effectively identify and reconcile contradictions between control objectives at different scales. It makes decisions based on the overall system performance, avoiding the problem of local optimization harming global performance, thus achieving true system-level overall optimization and improving overall operational efficiency.

[0017] 2. This invention introduces a closed-loop adaptive control and real-time monitoring feedback mechanism. The system can dynamically sense multi-scale state changes and control effects, and adjust the control strategy in a timely manner. This closed-loop process of "perception-decision-execution-feedback" enables the system to have stronger adaptive and recovery capabilities when facing changes in internal parameters, external disturbances, or local faults, significantly improving the system's robustness and overall stability.

[0018] 3. By combining machine learning algorithms and predictive control steps, this invention can not only uncover potential patterns from multi-scale historical data, but also predict the future state evolution of the system. This makes the control strategy no longer a simple passive response, but rather possesses a certain degree of foresight and intelligence, enabling advance planning and prevention, thereby better addressing uncertainties such as fluctuations in renewable energy and changes in production orders.

[0019] 4. Through scale priority management and intelligent distribution by the coordinator, this invention can prioritize the control needs of the most critical scales for system security and economic operation, even when resources (such as computing resources, energy, and communication bandwidth) are limited. This on-demand allocation and dynamic adjustment mechanism ensures that control resources are used effectively, avoids resource waste, and achieves efficient cross-scale collaborative configuration.

[0020] 5. The method proposed in this invention is a general framework, and its core ideas can be widely applied to the control of complex dynamic systems in various fields such as power systems, intelligent manufacturing, and smart cities. By adapting to multi-scale data sources and execution units in different application scenarios, this method can provide a unified and effective technical solution for solving the coordination and control problems of various large-scale systems. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0022] The embodiments of the present invention will be further described below with reference to the accompanying drawings: Example 1: As Figure 1 As shown, the present invention provides a multi-scale coordinated control method, comprising the following steps: (1) Obtain multi-scale data from the controlled system, wherein the multi-scale data includes system state information at at least two different scales, wherein the scale includes at least one of time scale and spatial scale, and each scale corresponds to a different data acquisition frequency or resolution; This ensures that the data covers different levels of the system, from macro to micro.

[0023] (2) Perform scale feature analysis on the multi-scale data to identify system behavior patterns and relationships at each scale, wherein the scale feature analysis includes data fusion, pattern extraction and anomaly detection; It helps to understand the dynamic characteristics of the system at different scales.

[0024] (3) Based on the results of the scale feature analysis, a multi-scale coordinated control strategy is generated. The control strategy includes a set of control instructions for different scales. When generating the control strategy, the interaction and priority between scales are considered to ensure the coordination of control actions. (4) Execute the multi-scale coordinated control strategy, distribute control commands to the corresponding execution units of the controlled system through the coordinator, and dynamically adjust the control actions during the execution process to ensure the consistency of control objectives at different scales; (5) Monitor the control effect in real time and feed back the monitoring results to steps (1) to (4) to achieve closed-loop adaptive control.

[0025] The multi-scale coordinated control method is applicable to complex dynamic systems. By coordinating control actions at different scales, it improves the overall performance and robustness of the system.

[0026] The multi-scale data in step (1) includes macro-scale data, meso-scale data and micro-scale data. Macro-scale data reflects the overall operating status of the system, meso-scale data reflects the status of subsystems or regions, and micro-scale data reflects the status of local components or real-time details. The data acquisition is achieved through at least one of sensor networks, historical databases or simulation.

[0027] This ensured the comprehensiveness and diversity of the data, providing a foundation for subsequent analysis.

[0028] The mesoscale feature analysis in step (2) includes: Normalization: Normalize multi-scale data to eliminate the effects of scale differences; Machine learning algorithms or statistical analysis: Using machine learning algorithms or statistical analysis methods to uncover potential patterns and trends in data at various scales; Assess the inter-scale coupling strength: Assess the inter-scale coupling strength and generate a scale correlation matrix for subsequent control strategy generation; The scale feature analysis can identify stability and vulnerability indicators of the system at multiple scales, helping to assess system risk.

[0029] In step (3), when generating the multi-scale coordinated control strategy, a hierarchical decision-making mechanism is adopted, including: Set long-term or overall control objectives at a macro scale; Develop medium-term or regional control plans at the meso-scale; Formulate short-term or localized control actions at the microscale; The control strategy is generated through an optimization algorithm to minimize inter-scale conflicts and maximize system efficiency, and the control commands include setpoint adjustment, mode switching, or parameter tuning.

[0030] In step (4), the coordinator uses an event-driven or time-triggered mechanism to distribute control commands, and the coordination process includes: Control resources are dynamically allocated based on scale priority. Detect and resolve control conflicts between scales, and ensure the coordination of control actions through arbitration logic; Delay compensation and fault tolerance are introduced into control execution to deal with communication or execution anomalies; The coordinator implements intelligent decision-making based on fuzzy logic or a rule engine.

[0031] The multi-scale data includes grid frequency, voltage level, load demand and renewable energy output, and the control objectives include frequency stability, voltage optimization and energy dispatch; the monitoring feedback in step (5) is linked with the grid dispatch center to achieve multi-timescale power generation and power consumption coordination.

[0032] When generating the control strategy in step (3), a predictive control step is also included: Based on historical data and real-time information, predict the state evolution of the system at multiple scales in the future; Adjust control strategies in advance based on forecast results to cope with potential disturbances or changes in demand; The predictive control adopts a rolling optimization method and is combined with the feedback mechanism in step (5) to improve control accuracy.

[0033] The real-time monitoring in step (5) includes performance index calculation, which includes response time, energy consumption, stability or efficiency; and the feedback adjustment adopts an adaptive learning algorithm to automatically update the control parameters according to system changes. The adaptive learning algorithm includes neural networks, genetic algorithms or reinforcement learning.

[0034] It also includes scale priority management steps: Priority weights are assigned to different scales, with the priorities dynamically adjusted based on system criticality, real-time requirements, or external events. In the generation and execution of control strategies, high-scale control tasks are given priority. The priority management is implemented through weighted voting or consensus algorithms to ensure that the core functions of the system are not affected when resources are limited.

[0035] The method is applied to an intelligent manufacturing system, where multi-scale data includes equipment status, production progress, material flow and energy consumption, and the control objectives include maximizing production efficiency, fault prevention and resource optimization; the coordinated execution in step (4) is integrated with the manufacturing execution system to achieve multi-scale collaborative control from the factory level to the equipment level.

Claims

1. A multi-scale coordinated control method, characterized in that, Includes the following steps: (1) Obtain multi-scale data from the controlled system, wherein the multi-scale data includes system state information at at least two different scales, wherein the scale includes at least one of time scale and spatial scale, and each scale corresponds to a different data acquisition frequency or resolution; (2) Perform scale feature analysis on the multi-scale data to identify system behavior patterns and relationships at each scale, wherein the scale feature analysis includes data fusion, pattern extraction and anomaly detection; (3) Based on the results of the scale feature analysis, a multi-scale coordinated control strategy is generated. The control strategy includes a set of control instructions for different scales, wherein the interaction and priority between scales are considered when generating the control strategy. (4) Execute the multi-scale coordinated control strategy, distribute control commands to the corresponding execution units of the controlled system through the coordinator, and dynamically adjust the control actions during the execution process to ensure the consistency of control objectives at different scales; (5) Monitor the control effect in real time and feed back the monitoring results to steps (1) to (4) to achieve closed-loop adaptive control.

2. The multi-scale coordinated control method according to claim 1, characterized in that, The multi-scale data in step (1) includes macro-scale data, meso-scale data and micro-scale data. Macro-scale data reflects the overall operating status of the system, meso-scale data reflects the status of subsystems or regions, and micro-scale data reflects the status of local components or real-time details. The data acquisition is achieved through at least one of sensor networks, historical databases or simulation.

3. The multi-scale coordinated control method according to claim 1, characterized in that, The mesoscale feature analysis in step (2) includes: Normalize multi-scale data to eliminate the impact of scale differences; Employ machine learning algorithms or statistical analysis methods to uncover potential patterns and trends in data at various scales; The strength of coupling between scales is evaluated, and a scale correlation matrix is ​​generated for subsequent control strategy generation. The scale feature analysis can identify the stability and vulnerability indicators of the system at multiple scales.

4. The multi-scale coordinated control method according to claim 1, characterized in that, In step (3), when generating the multi-scale coordinated control strategy, a hierarchical decision-making mechanism is adopted, including: Set long-term or overall control objectives at a macro scale; Develop medium-term or regional control plans at the meso-scale; Formulate short-term or localized control actions at the microscale; The control strategy is generated through an optimization algorithm to minimize inter-scale conflicts and maximize system efficiency, and the control commands include setpoint adjustment, mode switching, or parameter tuning.

5. The multi-scale coordinated control method according to claim 1, characterized in that, In step (4), the coordinator uses an event-driven or time-triggered mechanism to distribute control commands, and the coordination process includes: Control resources are dynamically allocated based on scale priority. Detect and resolve control conflicts between scales, and ensure the coordination of control actions through arbitration logic; Delay compensation and fault tolerance are introduced into control execution to deal with communication or execution anomalies; The coordinator implements intelligent decision-making based on fuzzy logic or a rule engine.

6. The multi-scale coordinated control method according to claim 1, characterized in that, The multi-scale data includes grid frequency, voltage level, load demand and renewable energy output, and the control objectives include frequency stability, voltage optimization and energy dispatch; the monitoring feedback in step (5) is linked with the grid dispatch center to achieve multi-timescale power generation and power consumption coordination.

7. The multi-scale coordinated control method according to claim 1, characterized in that, When generating the control strategy in step (3), a predictive control step is also included: Based on historical data and real-time information, predict the state evolution of the system at multiple scales in the future; Adjust control strategies in advance based on forecast results to cope with potential disturbances or changes in demand; The predictive control adopts a rolling optimization method and is combined with the feedback mechanism in step (5) to improve control accuracy.

8. The multi-scale coordinated control method according to claim 1, characterized in that, The real-time monitoring in step (5) includes performance index calculation, which includes response time, energy consumption, stability or efficiency; and the feedback adjustment adopts an adaptive learning algorithm to automatically update the control parameters according to system changes. The adaptive learning algorithm includes neural networks, genetic algorithms or reinforcement learning.

9. The multi-scale coordinated control method according to claim 1, characterized in that, It also includes scale priority management steps: Priority weights are assigned to different scales, with the priorities dynamically adjusted based on system criticality, real-time requirements, or external events. In the generation and execution of control strategies, high-scale control tasks are given priority. The priority management is implemented through weighted voting or consensus algorithms to ensure that the core functions of the system are not affected when resources are limited.

10. The multi-scale coordinated control method according to claim 1, characterized in that, The method is applied to an intelligent manufacturing system, where multi-scale data includes equipment status, production progress, material flow and energy consumption, and the control objectives include maximizing production efficiency, fault prevention and resource optimization; the coordinated execution in step (4) is integrated with the manufacturing execution system to achieve multi-scale collaborative control from the factory level to the equipment level.

Images