Intelligent control system supporting edge computing

Abstract

The invention discloses an intelligent control system supporting edge computing, which belongs to the technical field of edge computing and comprises a self-adaptive environment sensing module, an edge intelligent path planning module, a heterogeneous equipment collaborative decision module, an equipment control module, an edge autonomous fault-tolerant control module, a dynamic energy consumption optimization module and an edge optimization module. According to the method, environment data and task requirements are processed in parallel in an edge node cluster through a distributed cooperative computing mechanism, a path is generated based on a multi-objective optimization model, and balance between path quality and resource consumption is realized through an edge computing delay optimization formula and a cooperative efficiency evaluation model; the weight coefficient is dynamically adjusted to adapt to different scene requirements; multi-device cooperative path planning is supported, and conflicts among devices are avoided through safety radius constraint and coordinate point optimization; the path planning response time is shortened to hundreds of milliseconds, the path conflict rate is reduced, and the multi-device collaborative operation efficiency and safety are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of edge computing technology, specifically referring to an intelligent control system that supports edge computing. Background Technology

[0002] With the surge in the number of IoT devices, traditional cloud-based centralized control architectures are facing problems such as high latency, bandwidth bottlenecks, and data privacy.

[0003] However, existing edge computing intelligent control systems still have certain shortcomings. Existing technologies rely on centralized cloud processing for path planning tasks, resulting in high data transmission and processing latency, which cannot meet real-time requirements. Path planning algorithms lack multi-objective optimization capabilities, only considering a single factor and ignoring key indicators such as network energy consumption and task unloading latency, resulting in a high path conflict rate. In multi-device collaborative scenarios, it is impossible to dynamically optimize paths through safety radius constraints, significantly increasing the risk of collisions between devices and severely restricting the efficiency and safety of collaborative operations. The lack of device resource pools and task decomposition mechanisms makes it impossible to effectively identify the differences in capabilities of heterogeneous devices, leading to unreasonable task allocation. Therefore, an intelligent control system supporting edge computing is proposed. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent control system that supports edge computing, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an intelligent control system supporting edge computing, comprising an adaptive environment perception module, an edge intelligent path planning module, a heterogeneous device collaborative decision-making module, a device control module, an edge autonomous fault-tolerant control module, a dynamic energy consumption optimization module, and an edge optimization module;

[0006] The adaptive environment perception module performs real-time dynamic environment acquisition by fusing data from multiple sensors at the device end;

[0007] The edge intelligent path planning module receives environmental data, performs distributed collaborative computing at edge nodes, and generates an optimized path.

[0008] The heterogeneous device collaborative decision-making module performs real-time scheduling and arbitration on the regional edge server based on global tasks and local constraints, and issues collaborative decision-making instructions to the device control module.

[0009] The equipment control module receives instructions for path planning and collaborative decision-making, performs high-precision, low-latency motion and operation control locally on the equipment, and synchronizes real-time status data to the edge autonomous fault-tolerant control module and the dynamic energy consumption optimization module.

[0010] The edge autonomous fault-tolerant control module monitors the device status in real time, executes the fault-tolerant strategy locally when an anomaly is detected, and reports the degraded operation status to the heterogeneous device collaborative decision-making module.

[0011] The dynamic energy consumption optimization module analyzes the real-time energy consumption data of the equipment operation, optimizes the charging and discharging and power scheduling strategies at the edge, and feeds back the energy constraints to the edge intelligent path planning module and the heterogeneous device collaborative decision-making module.

[0012] The edge optimization module collects operational data for continuous optimization.

[0013] Preferably, the adaptive environment perception module is wirelessly connected to the edge intelligent path planning module, the edge intelligent path planning module is wirelessly connected to the device control module, the heterogeneous device collaborative decision-making module is wirelessly connected to the device control module, the device control module is wirelessly connected to the edge autonomous fault-tolerant control module and the dynamic energy consumption optimization module, the edge autonomous fault-tolerant control module is wirelessly connected to the heterogeneous device collaborative decision-making module, the dynamic energy consumption optimization module is wirelessly connected to the heterogeneous device collaborative decision-making module and the edge intelligent path planning module, and the edge optimization module is wirelessly connected to the adaptive environment perception module.

[0014] Preferably, the adaptive environment perception module deploys multiple sensors for real-time data acquisition, presets a high-precision time synchronization mechanism at the edge node level, aligns the timing of data from multiple sensors, handles timing differences in sensor data arrival, performs preliminary filtering on the raw sensor data to remove obvious noise and outliers, converts the outputs of different sensors into a unified data format, and automatically adjusts the sensor operating parameters and acquisition strategy according to changes in environmental conditions.

[0015] Preferably, the edge intelligent path planning module acquires environmental data, determines core optimization objectives based on system requirements, and performs latency optimization through data processing at edge nodes. Edge computing latency optimization is implemented as follows:

[0016] ,

[0017] In the formula, This represents the amount of latency optimization brought about by edge computing. D represents the central node processing time, D represents the data transmission volume, and R represents the network transmission rate. This represents the computational workload of the i-th edge node. represents the processing capacity of the i-th edge node, and n represents the number of edge nodes participating in collaborative computing.

[0018] Preferably, the edge intelligent path planning module establishes a distributed computing coordination mechanism in the edge node cluster, allocates computing task scope to each edge node, and evaluates the edge collaborative computing efficiency, as follows:

[0019] ,

[0020] In the formula, Indicates the efficiency of edge collaborative computing. Indicates the network transmission impact coefficient;

[0021] The optimization of multi-target edge paths is calculated based on latency optimization and edge collaborative computing efficiency, and is implemented as follows:

[0022] ,

[0023] In the formula, This represents the optimized multi-target edge path. Indicates the minimum safe distance. This represents the weighting coefficient.

[0024] Preferably, when the regional edge server starts, the heterogeneous device collaborative decision-making module initializes the device identification module, automatically scans and identifies all heterogeneous devices connected to the network, obtains basic information such as device type, performance parameters, and current status, establishes a device resource pool, presets a global task database, stores information on various task types, priorities, and deadlines that the current system needs to complete, forms a task set, performs multi-dimensional analysis on each task based on the global task database, decomposes complex tasks into subtasks that can be processed in parallel according to task characteristics, forms a task decomposition tree, evaluates the execution feasibility of each subtask, excludes subtasks that cannot be executed, and assigns an initial priority to each subtask.

[0025] Preferably, the heterogeneous device collaborative decision-making module performs collaborative decision-making among devices through distributed collaborative scheduling, implements a dynamic task allocation mechanism, assigns tasks to the most suitable device based on the device's current load and capacity, and executes a conflict detection and arbitration mechanism. When multiple devices compete for the same resource or task, arbitration is performed based on preset rules and priorities, and the device priority score is calculated as follows:

[0026] ,

[0027] In the formula, This represents the priority score of device i. Indicates the maximum computing power of the system. This represents the computing power of device i. Indicates the minimum communication bandwidth in the system. This represents the communication bandwidth of device i. This represents the maximum time constraint in the system. This represents the time constraint for device i. This represents the weighting coefficient for computational capability. Indicates the communication capability weighting coefficient. This represents the time constraint weight coefficient; based on the usage of equipment resources, the task allocation is dynamically adjusted, and the final decision result is transformed into a structured instruction set for distribution.

[0028] Preferably, during the device startup phase, the device control module loads a high-precision real-time operating system, configures a microsecond-level timer, and initializes all control hardware interfaces. It receives instructions through dual-channel edge network, performs multi-dimensional verification on the instructions, converts the received path coordinate sequence and collaborative decision instructions into a sequence of control parameters that can be executed locally on the device, and sends the instructions directly into the control execution queue based on the urgency of the instructions. Low-priority instructions enter the buffer queue. During the control execution process, sensor data is fused in real time to fine-tune the path. In each control cycle, key status data is collected synchronously, and the processed data is packaged into lightweight messages and transmitted through the high-priority channel of the edge network.

[0029] Preferably, the edge autonomous fault-tolerant control module acquires real-time monitoring device status, and based on historical device operating data and environmental characteristics, presets device safety thresholds to calculate the severity of the device status, thereby achieving the following:

[0030] ,

[0031] In the formula, Indicates the severity of equipment malfunction. This indicates the current equipment status parameter value. Indicates the equipment's reference state value. Indicates the device's safety threshold. It represents the environmental fluctuation coefficient, immediately marks the detected anomalies as severe, and records the anomaly occurrence timestamp, parameter deviation, and environmental state snapshot.

[0032] Preferably, the edge autonomous fault-tolerant control module calculates the fault-tolerant execution strength based on the device anomaly severity index, and implements it as follows:

[0033] ,

[0034] In the formula, S represents the fault-tolerant execution strength. The threshold for triggering an anomaly is represented by 'k', and the fault tolerance strength coefficient is represented by 'k'. Based on the anomaly level and type, the optimal strategy is matched from the fault tolerance strategy library. When multiple anomalies occur simultaneously, the strategies are executed in a preset priority order. The strategy instructions are sent through the hardware-level instruction channel. After the strategy is executed, it is immediately verified whether the anomaly has been mitigated. If it has not been mitigated, the secondary fault tolerance strategy is automatically triggered, and the degraded operation status is reported.

[0035] Preferably, the dynamic energy consumption optimization module receives high-precision energy consumption data streams transmitted by the device control module in real time through a dedicated high-priority communication channel on the edge node. It performs multi-dimensional energy consumption pattern recognition based on a historical energy consumption database, predicts energy consumption using environmental perception data and task scheduling information, determines peak and off-peak energy consumption periods through cluster analysis, and constructs a multi-objective optimization function based on energy consumption prediction and the current energy status:

[0036] ,

[0037] In the formula, Indicates energy loss. Indicates peak load. This indicates fluctuations in battery status. The system allocates dynamic power priority to devices and implements device-level power peak shaving: when the total load exceeds a preset threshold, it automatically reduces the power of non-critical devices. The optimization results generate a sequence of scheduling instructions in real time, and the optimization results are structured into an energy constraint report. The report is synchronized to the edge intelligent path planning module and the heterogeneous device collaborative decision-making module through a highly reliable edge network channel.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] 1. This invention utilizes a distributed collaborative computing mechanism to process environmental data and task requirements in parallel within an edge node cluster, overcoming the latency bottleneck of traditional centralized path planning. It generates paths based on a multi-objective optimization model, achieving a balance between path quality and resource consumption through edge computing latency optimization formulas and collaborative efficiency evaluation models. It dynamically adjusts weight coefficients to adapt to different scenario requirements; supports multi-device collaborative path planning; and avoids inter-device conflicts through safety radius constraints and coordinate point optimization. It reduces path planning response time to the hundreds of milliseconds level, lowers path conflict rates, and significantly improves the efficiency and safety of multi-device collaborative operations.

[0040] 2. This invention achieves efficient decomposition and dynamic scheduling of complex tasks through a device resource pool and task decomposition tree mechanism; it constructs a multi-dimensional decision-making model based on device capability matrix and task priority to solve the problem of heterogeneous device resource allocation; it introduces dynamic weighted device priority scoring, comprehensively considering indicators such as computing power, communication bandwidth, and time constraints to ensure the fairness and efficiency of task allocation; it handles resource competition issues in real time through conflict detection and arbitration mechanisms to avoid task blocking; it improves device resource utilization, shortens task completion time, and is suitable for industrial scenarios with diverse device types and high task concurrency, significantly reducing system operating costs.

[0041] 3. This invention significantly improves system reliability through local anomaly detection and dynamic fault tolerance strategies. Based on an anomaly severity quantification model using environmental fluctuation coefficients, it achieves adaptive threshold adjustment, reducing false alarm rates. The hierarchical fault tolerance strategy library supports adjustments from mild to severe interventions, ensuring that anomaly handling matches system load. The fault tolerance strategy is executed through a hardware-level instruction channel, shortening system recovery time and reducing equipment downtime in sudden failure scenarios, thus providing safety assurance for high-risk industrial environments. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the structure of an intelligent control system supporting edge computing according to the present invention;

[0043] Figure 2 The present invention describes the operation flow of an intelligent control system supporting edge computing. Figure 1 ;

[0044] Figure 3 The present invention describes the operation flow of an intelligent control system supporting edge computing. Figure 2 ;

[0045] Figure 4 The present invention describes the operation flow of an intelligent control system supporting edge computing. Figure 3 . Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Example

[0048] Please see Figures 1-4As shown, the present invention provides a technical solution including an adaptive environment perception module, an edge intelligent path planning module, a heterogeneous device collaborative decision-making module, a device control module, an edge autonomous fault-tolerant control module, a dynamic energy consumption optimization module, and an edge optimization module;

[0049] The adaptive environment perception module performs real-time dynamic environment acquisition by fusing data from multiple sensors at the device end;

[0050] The edge intelligent path planning module receives environmental data, performs distributed collaborative computing at edge nodes, and generates an optimized path.

[0051] The heterogeneous device collaborative decision-making module performs real-time scheduling and arbitration on the regional edge server based on global tasks and local constraints, and issues collaborative decision-making instructions to the device control module.

[0052] The equipment control module receives instructions for path planning and collaborative decision-making, performs high-precision, low-latency motion and operation control locally on the equipment, and synchronizes real-time status data to the edge autonomous fault-tolerant control module and the dynamic energy consumption optimization module.

[0053] The edge autonomous fault-tolerant control module monitors the device status in real time, executes the fault-tolerant strategy locally when an anomaly is detected, and reports the degraded operation status to the heterogeneous device collaborative decision-making module.

[0054] The dynamic energy consumption optimization module analyzes the real-time energy consumption data of the equipment operation, optimizes the charging and discharging and power scheduling strategies at the edge, and feeds back the energy constraints to the edge intelligent path planning module and the heterogeneous device collaborative decision-making module.

[0055] The edge optimization module collects operational data for continuous optimization.

[0056] In this embodiment, the adaptive environment perception module is wirelessly connected to the edge intelligent path planning module, the edge intelligent path planning module is wirelessly connected to the device control module, the heterogeneous device collaborative decision-making module is wirelessly connected to the device control module, the device control module is wirelessly connected to the edge autonomous fault-tolerant control module and the dynamic energy consumption optimization module, the edge autonomous fault-tolerant control module is wirelessly connected to the heterogeneous device collaborative decision-making module, the dynamic energy consumption optimization module is wirelessly connected to the heterogeneous device collaborative decision-making module and the edge intelligent path planning module, and the edge optimization module is wirelessly connected to the adaptive environment perception module.

[0057] In this embodiment, the adaptive environment perception module deploys multiple sensors for real-time data acquisition, presets a high-precision time synchronization mechanism at the edge node level, aligns the timing of data from multiple sensors, handles timing differences in sensor data arrival, performs preliminary filtering on the raw sensor data to remove obvious noise and outliers, converts the outputs of different sensors into a unified data format, and automatically adjusts the sensor operating parameters and acquisition strategies according to changes in environmental conditions.

[0058] In this embodiment, the edge intelligent path planning module acquires environmental data and determines core optimization objectives based on system requirements, including task offloading latency, network energy consumption, load balancing, and flight safety. Edge computing reduces latency and performs latency optimization by processing data at edge nodes. Edge computing latency optimization is implemented as follows:

[0059] ,

[0060] In the formula, This represents the amount of latency optimization brought about by edge computing. D represents the central node processing time, D represents the data transmission volume, and R represents the network transmission rate. This represents the computational workload of the i-th edge node. represents the processing capacity of the i-th edge node, and n represents the number of edge nodes participating in collaborative computing.

[0061] In this embodiment, the edge intelligent path planning module establishes a distributed computing coordination mechanism in the edge node cluster, allocates computing task scope to each edge node, and evaluates the edge collaborative computing efficiency, which is implemented as follows:

[0062] ,

[0063] In the formula, Indicates the efficiency of edge collaborative computing. The larger the value, the higher the efficiency of edge collaborative computing. Indicates the network transmission impact coefficient;

[0064] Specifically, the optimization of multi-target edge paths is calculated based on latency optimization and edge collaborative computing efficiency, as follows:

[0065] ,

[0066] In the formula, This represents the optimized multi-target edge path. The smaller the value, the higher the quality of path planning. Indicates the minimum safe distance. for , This represents the coordinates of the i-th path. This represents the safe radius of the i-th path. This represents the weighting coefficient.

[0067] In this embodiment, when the regional edge server starts, the heterogeneous device collaborative decision-making module initializes the device identification module, automatically scans and identifies all heterogeneous devices connected to the network, obtains basic information such as device type, performance parameters, and current status, establishes a device resource pool, records key indicators such as computing power, storage capacity, communication bandwidth, and energy consumption characteristics of each device, and constructs a device capability matrix.

[0068] Specifically, a global task database is pre-set to store information such as the types, priorities, and deadlines of various tasks that the current system needs to complete, forming a task set. Based on the global task database, each task is analyzed from multiple dimensions, including task type, required resources, time constraints, and priority. According to the task characteristics, complex tasks are decomposed into subtasks that can be processed in parallel, forming a task decomposition tree. The execution feasibility of each subtask is evaluated, taking into account the device capability matrix and environmental constraints, eliminating subtasks that cannot be executed, and assigning an initial priority to each subtask, taking into account factors such as the urgency of the task and its impact on the global goal.

[0069] In this embodiment, the heterogeneous device collaborative decision-making module performs collaborative decision-making among devices through distributed collaborative scheduling, implements a dynamic task allocation mechanism, allocates tasks to the most suitable device based on the current load and capacity of the device, and executes a conflict detection and arbitration mechanism.

[0070] Specifically, when multiple devices compete for the same resource or task, arbitration is performed based on preset rules and priorities, and the device priority score is calculated as follows:

[0071] ,

[0072] In the formula, This represents the priority score of device i. A higher value indicates that device i has a greater advantage in arbitration. Indicates the maximum computing power of the system. This represents the computing power of device i. Indicates the minimum communication bandwidth in the system. This represents the communication bandwidth of device i. This represents the maximum time constraint in the system. This represents the time constraint for device i. This represents the weighting coefficient for computational capability. Indicates the communication capability weighting coefficient. This represents the time constraint weighting coefficient.

[0073] Based on the usage of equipment resources, the task allocation is dynamically adjusted, and the final decision result is transformed into a structured instruction set for distribution, including task allocation information, execution parameters, time requirements, etc.

[0074] In this embodiment, during the device startup phase, the device control module loads a high-precision real-time operating system, configures a microsecond-level timer, initializes all control hardware interfaces, receives instructions through a dual-channel edge network, performs multi-dimensional verification on the instructions, including syntax parsing, device capability matching, and timestamp alignment, and converts the received path coordinate sequence and collaborative decision instructions into a sequence of control parameters that can be executed locally on the device.

[0075] Specifically, based on the urgency of the command, the command is directly sent to the control execution queue, while low-priority commands enter the buffer queue. During the control execution process, sensor data is fused in real time to fine-tune the path. In each control cycle, key status data is collected synchronously, and the processed data is packaged into lightweight messages and transmitted through the high-priority channel of the edge network.

[0076] In this embodiment, the edge autonomous fault-tolerant control module acquires real-time monitoring device status, and calculates the severity of device status based on historical device operating data and environmental characteristics, presets device security thresholds, and performs the following:

[0077] ,

[0078] In the formula, This index indicates the severity of the equipment malfunction; a higher value indicates a more severe malfunction. This indicates the current equipment status parameter value. Indicates the equipment's reference state value. Indicates the device's safety threshold. It represents the environmental fluctuation coefficient, immediately marks the detected anomalies as severe, and records the anomaly occurrence timestamp, parameter deviation, and environmental state snapshot.

[0079] In this embodiment, the edge autonomous fault-tolerant control module calculates the fault-tolerant execution strength based on the device anomaly severity index, and implements it as follows:

[0080] ,

[0081] In the formula, S represents the fault tolerance execution strength; a larger value indicates a more aggressive fault tolerance strategy. This represents the abnormal trigger threshold, and k represents the fault tolerance strength coefficient.

[0082] Specifically, based on the level and type of the exception, the optimal strategy is matched from the fault tolerance strategy library. When multiple exceptions occur simultaneously, the strategies are executed in a preset priority order. The strategy instructions are sent through the hardware-level instruction channel. After the strategy is executed, it is immediately verified whether the exception has been mitigated. If it has not been mitigated, the secondary fault tolerance strategy is automatically triggered and the downgraded operation status is reported.

[0083] In this embodiment, the dynamic energy consumption optimization module receives high-precision energy consumption data streams transmitted by the device control module in real time through a dedicated high-priority communication channel for edge nodes. Based on a historical energy consumption database, it performs multi-dimensional energy consumption pattern recognition using an improved LSTM neural network model, predicts energy consumption using environmental perception data and task scheduling information, determines peak and off-peak energy consumption periods through cluster analysis, and constructs a multi-objective optimization function based on energy consumption prediction and the current energy status:

[0084] ,

[0085] In the formula, Indicates energy loss. Indicates peak load. This indicates fluctuations in battery status. .

[0086] Specifically, dynamic power priority is allocated to devices, and device-level power peak shaving is implemented: when the total load exceeds the preset threshold, the power of non-critical devices is automatically reduced, the optimization results are used to generate a sequence of scheduling instructions in real time, the optimization results are structured into an energy constraint report, and the report is synchronized to the edge intelligent path planning module and the heterogeneous device collaborative decision-making module through a highly reliable edge network channel.

[0087] Working principle: The system collects environmental data in real time through multiple sensors deployed on the device and uses the high-precision time synchronization mechanism of edge nodes to solve the problem of data timing differences between sensors. In the data preprocessing stage, the system removes noise and outliers through filtering algorithms, unifies the output format of heterogeneous sensors, and forms a standardized environmental state model. According to real-time environmental changes, the module dynamically adjusts the sampling frequency and operating parameters of the sensors.

[0088] Based on the distributed collaborative computing capabilities of edge node clusters, the system receives real-time data from the environmental perception module and constructs a multi-objective optimization model in conjunction with system requirements. Edge computing reduces reliance on central servers, shortening data transmission and processing latency. Within edge nodes, the system divides the scope of computing tasks, evaluates collaborative computing efficiency, and generates optimized paths based on network transmission impact coefficients. In path planning, the system comprehensively considers minimum safe distance, device coordinate distribution, and path risk weights. A device resource pool is initialized in the regional edge server, automatically identifying and classifying heterogeneous devices connected to the network, recording key indicators such as computing power, communication bandwidth, and energy consumption characteristics. Through a global task database, the system analyzes task types, priorities, and constraints, decomposing complex tasks into parallelizable subtasks and excluding inexecutable subtasks. A dynamic task allocation strategy is adopted to match the optimal executor based on the device's current load and capabilities, and resolves resource contention issues through conflict detection and arbitration mechanisms. Finally, the system generates structured collaborative decision-making instructions. A high-precision real-time operating system runs locally on the devices, receiving path planning and collaborative decision-making instructions through dual-channel edge network. After parsing the instructions, matching device capabilities, and aligning timestamps, the system converts them into a sequence of executable control parameters. High-priority instructions... Commands are directly entered into the execution queue, while low-priority commands are buffered and wait. Within the control cycle, the module integrates sensor data to fine-tune the path in real time, and the system synchronously collects equipment status data, transmitting it to the fault tolerance and energy consumption optimization module through a high-priority channel. It continuously monitors the equipment's operating status and sets dynamic safety thresholds based on historical data and environmental characteristics. When an anomaly is detected, the system calculates the severity of the anomaly and marks its level, recording the time of occurrence and environmental conditions. The module selects a matching processing solution from the fault tolerance strategy library based on the anomaly type and level, and executes it quickly through the hardware-level command channel. If the initial strategy fails to mitigate the anomaly, the system triggers a secondary fault tolerance mechanism. The system reports equipment degradation status to the collaborative decision-making module; receives equipment energy consumption data in real time through a dedicated communication channel, and predicts future energy consumption trends by combining historical energy consumption pattern recognition and environmental perception information; the system determines peak and off-peak energy consumption periods through cluster analysis and formulates charging, discharging, and power scheduling strategies based on a multi-objective optimization model; when the total load exceeds a threshold, the module dynamically adjusts the power allocation of non-critical equipment, generates an energy constraint report, and feeds it back to the path planning and collaborative decision-making module; it continuously collects system operation data, identifies performance bottlenecks and optimization space through data analysis; and the system dynamically adjusts the parameter configuration and strategy rules of each module based on historical data.

[0089] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their likenesses.

[0090] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An intelligent control system supporting edge computing, characterized in that: It includes an adaptive environment perception module, an edge intelligent path planning module, a heterogeneous device collaborative decision-making module, a device control module, an edge autonomous fault-tolerant control module, a dynamic energy consumption optimization module, and an edge optimization module; The adaptive environment perception module performs real-time dynamic environment acquisition by fusing data from multiple sensors at the device end; The edge intelligent path planning module receives environmental data, performs distributed collaborative computing at edge nodes, and generates an optimized path. The heterogeneous device collaborative decision-making module performs real-time scheduling and arbitration on the regional edge server based on global tasks and local constraints, and issues collaborative decision-making instructions to the device control module. The equipment control module receives instructions for path planning and collaborative decision-making, performs high-precision, low-latency motion and operation control locally on the equipment, and synchronizes real-time status data to the edge autonomous fault-tolerant control module and the dynamic energy consumption optimization module. The edge autonomous fault-tolerant control module monitors the device status in real time, executes the fault-tolerant strategy locally when an anomaly is detected, and reports the degraded operation status to the heterogeneous device collaborative decision-making module. The dynamic energy consumption optimization module analyzes the real-time energy consumption data of the equipment operation, optimizes the charging and discharging and power scheduling strategies at the edge, and feeds back the energy constraints to the edge intelligent path planning module and the heterogeneous device collaborative decision-making module. The edge optimization module collects operational data for continuous optimization.

2. The intelligent control system supporting edge computing according to claim 1, characterized in that: The adaptive environment perception module deploys multiple sensors for real-time data acquisition, presets a high-precision time synchronization mechanism at the edge node level, aligns the timing of data from multiple sensors, handles timing differences in sensor data arrival, performs preliminary filtering on the raw sensor data to remove obvious noise and outliers, converts the outputs of different sensors into a unified data format, and automatically adjusts sensor operating parameters and acquisition strategies according to changes in environmental conditions.

3. The intelligent control system supporting edge computing according to claim 1, characterized in that: The edge intelligent path planning module acquires environmental data and determines the core optimization objectives based on system requirements. Edge computing performs latency optimization by processing data at edge nodes. The edge computing latency optimization is implemented as follows: ， In the formula, This represents the amount of latency optimization brought about by edge computing. D represents the central node processing time, D represents the data transmission volume, and R represents the network transmission rate. This represents the computational workload of the i-th edge node. represents the processing capacity of the i-th edge node, and n represents the number of edge nodes participating in collaborative computing.

4. The intelligent control system supporting edge computing according to claim 3, characterized in that: The edge intelligent path planning module establishes a distributed computing coordination mechanism in the edge node cluster, allocates computing task scope to each edge node, and evaluates the efficiency of edge collaborative computing, as follows: ， In the formula, Indicates the efficiency of edge collaborative computing. Indicates the network transmission impact coefficient; The optimization of multi-target edge paths is calculated based on latency optimization and edge collaborative computing efficiency, and is implemented as follows: ， In the formula, This represents the optimized multi-target edge path. Indicates the minimum safe distance. This represents the weighting coefficient.

5. The intelligent control system supporting edge computing according to claim 4, characterized in that: The heterogeneous device collaborative decision-making module initializes the device identification module when the regional edge server starts, automatically scans and identifies all heterogeneous devices connected to the network, obtains basic information such as device type, performance parameters, and current status, establishes a device resource pool, presets a global task database, stores information on various task types, priorities, and deadlines that the current system needs to complete, forms a task set, performs multi-dimensional analysis on each task based on the global task database, decomposes complex tasks into subtasks that can be processed in parallel according to task characteristics, forms a task decomposition tree, evaluates the execution feasibility of each subtask, excludes subtasks that cannot be executed, and assigns an initial priority to each subtask.

6. The intelligent control system supporting edge computing according to claim 5, characterized in that: The heterogeneous device collaborative decision-making module performs collaborative decision-making among devices through distributed collaborative scheduling, implements a dynamic task allocation mechanism, and assigns tasks to the most suitable device based on the current load and capacity of the device. It also executes a conflict detection and arbitration mechanism; when multiple devices compete for the same resource or task, arbitration is performed based on preset rules and priorities, and the device priority score is calculated as follows: ， In the formula, This represents the priority score of device i. Indicates the maximum computing power of the system. This represents the computing power of device i. Indicates the minimum communication bandwidth in the system. This represents the communication bandwidth of device i. This represents the maximum time constraint in the system. This represents the time constraint for device i. This represents the weighting coefficient for computational capability. Indicates the communication capability weighting coefficient. This represents the time constraint weighting coefficient; based on the usage of equipment resources, the task allocation is dynamically adjusted, and the final decision result is transformed into a structured instruction set for distribution.

7. The intelligent control system supporting edge computing according to claim 1, characterized in that: During the device startup phase, the device control module loads a high-precision real-time operating system, configures a microsecond-level timer, and initializes all control hardware interfaces. It receives instructions through dual-channel edge network, performs multi-dimensional verification on the instructions, and converts the received path coordinate sequence and collaborative decision instructions into a sequence of control parameters that can be executed locally on the device. Based on the urgency of the instructions, the instructions are directly sent to the control execution queue, while low-priority instructions are sent to the buffer queue. During the control execution process, sensor data is fused in real time to fine-tune the path. In each control cycle, key status data is collected synchronously, and the processed data is packaged into lightweight messages and transmitted through the high-priority channel of the edge network.

8. The intelligent control system supporting edge computing according to claim 1, characterized in that: The edge autonomous fault-tolerant control module acquires real-time monitoring device status, and based on historical device operating data and environmental characteristics, presets device security thresholds to calculate the severity of device status, which is implemented as follows: ， In the formula, Indicates the severity of equipment malfunction. This indicates the current equipment status parameter value. Indicates the equipment's reference state value. Indicates the device's safety threshold. It represents the environmental fluctuation coefficient, immediately marks the detected anomalies as severe, and records the anomaly occurrence timestamp, parameter deviation, and environmental state snapshot.

9. The intelligent control system supporting edge computing according to claim 1, characterized in that: The edge autonomous fault-tolerant control module calculates the fault-tolerant execution strength based on the device anomaly severity index, and implements it as follows: ， In the formula, S represents the fault-tolerant execution strength. The threshold for triggering an anomaly is represented by 'k', and the fault tolerance strength coefficient is represented by 'k'. Based on the anomaly level and type, the optimal strategy is matched from the fault tolerance strategy library. When multiple anomalies occur simultaneously, the strategies are executed in a preset priority order. The strategy instructions are sent through the hardware-level instruction channel. After the strategy is executed, it is immediately verified whether the anomaly has been mitigated. If it has not been mitigated, the secondary fault tolerance strategy is automatically triggered, and the degraded operation status is reported.

10. The intelligent control system supporting edge computing according to claim 1, characterized in that: The dynamic energy consumption optimization module receives high-precision energy consumption data streams transmitted by the device control module in real time through a dedicated high-priority communication channel on the edge node. It performs multi-dimensional energy consumption pattern recognition based on a historical energy consumption database, predicts energy consumption using environmental perception data and task scheduling information, determines peak and off-peak energy consumption periods through cluster analysis, and constructs a multi-objective optimization function based on energy consumption prediction and the current energy status: ， In the formula, Indicates energy loss. Indicates peak load. This indicates fluctuations in battery status. The system allocates dynamic power priority to devices and implements device-level power peak shaving: when the total load exceeds a preset threshold, it automatically reduces the power of non-critical devices. The optimization results generate a sequence of scheduling instructions in real time, and the optimization results are structured into an energy constraint report. The report is synchronized to the edge intelligent path planning module and the heterogeneous device collaborative decision-making module through a highly reliable edge network channel.

Images