Real-time monitoring system integration comprehensive management control system

Abstract

The invention discloses a real-time monitoring system integration comprehensive management control system comprising a system access module used for accessing lighting, air conditioning, electric power, security and protection and access control subsystems and collecting state data; the data processing module is used for the edge node to generate a state data set and a task queue; the modeling module is used for constructing a task set, an equipment set and a task-equipment mapping graph; the scheduling module is used for performing task allocation and scheduling variable updating based on an improved MISATPS algorithm; the priority control module is used for generating a priority scheduling queue; the consensus module is used for generating a consistency scheduling result based on a weighted Byzantine fault-tolerant consensus algorithm; the instruction execution module is used for issuing a control instruction and monitoring an execution state; and the rescheduling module is used for triggering a rescheduling process and generating an updating instruction under an abnormal condition. According to the invention, multi-subsystem cooperative control and high-reliability task scheduling are realized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent system integration and control technology, and in particular to a system integration comprehensive management and control system for real-time monitoring. Background Art

[0002] With the development of the Internet of Things, edge computing, and multi-system collaborative control technologies, buildings, industrial parks, and city-level smart facilities are gradually evolving towards distributed intelligence. Lighting systems, air-conditioning systems, power systems, security systems, and access control systems, as basic operating units, are increasingly required to be connected to a unified management platform to achieve centralized scheduling and intelligent joint control. To this end, system integration control has become a key direction of technological development. Its goal is to achieve cross-system status fusion and task control linkage by building a unified data perception, scheduling, and feedback control framework.

[0003] In the existing technology, the more common system integration control architecture mostly adopts a centralized management mode, that is, each subsystem is connected to the central main control platform through a communication interface, and various status data and control feedback information are uniformly collected. The main control platform centrally schedules tasks and issues control instructions. This centralized control method has a simple structure and clear logic, but in high-frequency tasks and multi-node equipment environments, there are problems such as large response delays, concentrated scheduling pressure, and serious communication bottlenecks. It is difficult to meet the engineering requirements of multi-task parallelism, high-availability control and fault isolation.

[0004] To overcome the bottleneck of the centralized processing mode, some technical solutions introduce edge computing architecture. By deploying edge nodes at the network boundary close to the system access module, they can locally process the collected operating status data and control feedback, build a preliminary status data set, and thus reduce the data load of the main control platform. However, the task processing of existing edge nodes still relies on rule-driven, static configuration or simple heuristic strategies, and lacks the scheduling and adaptive capabilities for status data sets and real-time resource status. At the same time, there is usually a lack of a unified task mapping model and scheduling variable set management mechanism between different edge nodes, which can easily lead to task allocation conflicts, equipment overload or resource idling.

[0005] Most current task scheduling methods do not perform structured modeling of the mapping relationship between task sets and device sets. The definition and evolution of scheduling variable sets lack controllability. The objective functions in the scheduling process are mostly single-objective or static weighted models, failing to integrate multiple system performance indicators such as response delay, resource energy consumption, load balancing, and collaborative coverage, making it difficult to meet multi-objective optimization requirements. At the same time, in the control feedback process, there is a lack of linkage mechanism between task execution results and scheduling variable sets, and feedback-driven scheduling adjustments cannot be achieved according to changes in the state data set, resulting in the system lacking adaptive scheduling capabilities.

[0006] In order to improve the efficiency and structural adaptability of multi-task scheduling, some studies have proposed a two-stage scheduling framework, among which the MISATPS algorithm is a typical representative. This algorithm optimizes the scheduling process in stages, taking into account scheduling efficiency and decoupling control, and has been applied in process control and scheduling optimization. However, the original MISATPS algorithm still has the following core defects in system integration control applications: it uses a fixed scheduling window and cannot adjust the scheduling granularity when the density of the task set changes or the resources of the equipment set fluctuate. It lacks a response mechanism to the system status. The algorithm does not consider the temporal evolution trend in the status data set and lacks short-term prediction function for the task status, resulting in a delayed response of the scheduling model when facing dynamically changing tasks. When some scheduling fails or resource conflicts occur, the original MISATPS algorithm usually needs to re-execute the full task allocation, and cannot achieve local optimization correction of the current scheduling variable set, which limits the real-time and robustness of the system.

[0007] In addition, the fault-tolerant mechanism design for the coordinated scheduling of multiple edge nodes in the current system is still not perfect. In the absence of a consensus control mechanism, different edge nodes may generate different sets of scheduling variables for the same set of tasks, causing the control execution unit to receive conflicting control instructions and cause operational abnormalities. Some studies have attempted to introduce Byzantine fault-tolerant algorithms into the consensus process, but they often ignore the differences between edge nodes in task execution accuracy, communication stability, and data integrity. The use of a unified weight voting strategy cannot effectively isolate the interference of abnormal nodes, nor can it dynamically adjust the consensus threshold to adapt to network fluctuations.

[0008] At the scheduling and execution level, current control execution units mostly execute tasks through control instructions issued by the main control platform. Some systems have the ability to return status data, but lack an abnormal feedback judgment mechanism and task-level rescheduling trigger logic. Scheduling failure usually means that the entire task set needs to be reinitialized and allocated, which has high computational overhead and long recovery time. It cannot support local disturbance correction of abnormal tasks, nor can it maintain the continuity and stability of the original scheduling structure.

[0009] In summary, the existing technology has obvious deficiencies in the integrated modeling of system access modules, the state data set processing of edge nodes, the mapping construction between task sets and device sets, the adaptive evolution of scheduling variable sets, the multi-objective construction of scheduling evaluation models, the weight adjustment and trusted voting mechanism of the consensus mechanism, and the anomaly detection and micro-disturbance rescheduling strategy of the control execution unit.

[0010] Therefore, how to provide a real-time monitoring system integrated comprehensive management and control system is a problem that those skilled in the art urgently need to solve. Summary of the Invention

[0011] One purpose of the present invention is to propose a real-time monitoring system integrated management and control system. The present invention constructs a unified modeling structure between a system access module, a state data set, a task set, a device set and a scheduling variable set, and adopts an improved MISAPS algorithm that combines a sliding scheduling window and a prediction feedback mechanism to complete the dynamic allocation of the task set. An optimization objective function including response delay, resource energy consumption, load balancing and coordination is set to construct a scheduling evaluation model. The scheduling order is adjusted through a priority scoring mechanism. A weighted Byzantine fault-tolerant consensus algorithm is used to perform consistency judgment on the scheduling variable set, and a control instruction is generated from the consistent scheduling result and sent to the control execution unit. When the state feedback is abnormal, a micro-disturbance rescheduling process is automatically triggered. The present invention has the advantages of high scheduling efficiency, fast control response, strong stability of scheduling results and high system fault tolerance and robustness.

[0012] A system integrated management and control system for real-time monitoring according to an embodiment of the present invention includes:

[0013] System access module, used to access lighting, air conditioning, power, security and access control subsystems, configure communication interfaces and control execution units, establish communication connections with the main control platform through edge nodes, and collect operating status data and control feedback;

[0014] The data processing module is used to perform denoising, alignment and caching operations on the collected data by the edge nodes to generate status data sets and task queues;

[0015] Modeling module, used to build task sets and equipment sets, generate task-equipment mapping graphs, and initialize scheduling variable sets;

[0016] The scheduling module is used to set the optimization objective function, build a scheduling evaluation model, and allocate the task set and update the scheduling variable set based on the improved MISAPTS algorithm combined with the sliding scheduling window and prediction feedback mechanism;

[0017] The priority control module is used to generate a priority scheduling queue based on the task urgency, device status and communication delay score;

[0018] The consensus module is used to execute the weighted Byzantine fault-tolerant consensus algorithm when the number of edge nodes is not less than two, and generate consistent scheduling results for the scheduling variable set;

[0019] The instruction execution module is used by the main control platform to generate control instruction sequences and send them to the control execution unit, and receive feedback status data for execution monitoring;

[0020] The rescheduling module is used to trigger the rescheduling process based on the current scheduling variable set under abnormal status or interrupt conditions and regenerate control instructions.

[0021] Optionally, modules can be connected using the following methods:

[0022] S1. Build a system integration architecture, deploy system access modules, and establish a communication connection with the main control platform through edge nodes;

[0023] S2, collects the operating status data and control feedback of the system access module, and generates a status data set and task queue after processing by the edge node;

[0024] S3. Build a task set and a device set based on the state data set and the task queue, generate a task-device mapping graph, and initialize a scheduling variable set;

[0025] S4. Set the optimization objective function and build a scheduling evaluation model by combining the task-equipment mapping diagram and the scheduling variable set;

[0026] S5. Introduce the improved MISAPTS algorithm, combine the sliding scheduling window with the prediction feedback mechanism, allocate the task set and update the scheduling variable set according to the scheduling evaluation model;

[0027] S6. Build a priority scoring mechanism to adjust the scheduling order based on the urgency of the task set, the status of the device set, and communication delay;

[0028] S7. When the number of edge nodes is not less than two, a weighted Byzantine fault-tolerant consensus algorithm is used to perform consensus calculation on the scheduling variable set to generate a consistent scheduling result.

[0029] S8. Convert the consistency scheduling result into a control instruction, which is sent by the main control platform to the control execution unit in the device set. When an execution exception occurs, the rescheduling process is triggered.

[0030] Optionally, the system access module includes a lighting system access unit, an air-conditioning system access unit, a power system access unit, a security system access unit and an access control system access unit. Each access unit is configured with a communication interface and a control execution unit, and establishes a data communication connection with the main control platform through an edge node for collecting operating status data, environmental parameters and control feedback signals.

[0031] Optionally, the improved MISAPTS algorithm includes:

[0032] A sliding scheduling window mechanism is introduced to dynamically update the length and stride of the sliding scheduling window based on the execution delay ratio of the task set, the resource utilization rate of the device set, and the adjustment deviation of the scheduling variable set in the previous scheduling cycle;

[0033] Embed a prediction feedback mechanism to predict the status of upcoming tasks in the task set based on the short-term change trend of the status dataset, and embed the prediction results as correction factors in the scheduling evaluation model;

[0034] When the control execution unit fails or there is an abnormal state in the device set, the perturbation rescheduling mechanism is triggered, the scheduling variable set is locally adjusted, and the task set is redistributed only within the corresponding device neighborhood, leaving the rest of the scheduling structure unchanged.

[0035] Optionally, the S3 specifically includes:

[0036] S31, extracting task information based on the status data set and constructing a task set, where each task in the task set includes a task identifier, resource requirements, an estimated duration, and task dependencies;

[0037] S32. Construct a device set based on the device information of the system access module, where each device in the device set includes a device identifier, processing performance, current status, and communication delay parameters;

[0038] S33. Generate a task-device mapping graph based on the adaptation relationship between the task set and the device set. The task-device mapping graph is a bipartite graph structure. The existence of an edge indicates that the task can be executed on the corresponding device.

[0039] S34, initialize the scheduling variable set, the scheduling variable includes the task identifier T i , Equipment Identification D j , scheduling relationship variable x ij , where x ij =1 indicates task T i Assigned to device D j , otherwise x ij =0.

[0040] Optionally, the S4 specifically includes:

[0041] S41. Set the scheduling optimization objective function:

[0042]

[0043] Among them, F is the scheduling optimization objective function value, λ1, λ2, λ3, λ4 are weight coefficients, n is the number of tasks in the task set, m is the number of devices in the device set, ω i For task T in the task set i The urgency factor, D i For task T i The response delay, θ j For device D in the device collection j The unit energy consumption coefficient, E j For device D j Energy consumption, Var(U j ) represents the device load value U in the device set jThe variance of |Δ| is the number of task pairs that meet the coordination constraints in the completed task set, and |Φ| is the total number of task pairs that need to be coordinated in the task set;

[0044] S42, constructing a scheduling evaluation model based on the task-device mapping diagram and the scheduling variable set, the scheduling evaluation model is based on the task identifier T in the task set i , device identifier D in the device collection j With the scheduling variable x ij As input, the optimization objective function F is used to represent the scoring result of the current scheduling scheme;

[0045] S43. The scheduling evaluation model is used to evaluate the allocation scheme between the task set and the equipment set during the scheduling process, and serves as a basis for task scheduling decisions.

[0046] Optionally, the S5 specifically includes:

[0047] S51. Calculate an initial scheduling score for each task in the task set according to the scheduling evaluation model, call the improved MISAPTS algorithm, combine the task-device mapping graph with the scheduling variable set, initialize the first round of task allocation plan and generate the scheduling variable set;

[0048] S52. Introduce a sliding scheduling window mechanism in the scheduling process, set the initial window length and window step, select only a subset of tasks within the window range for allocation in each round of scheduling, and record the scheduling feedback index E. t Then execute window update:

[0049]

[0050] Among them, W t+1 is the length of the next sliding window, W t is the current window length, η is the adjustment coefficient, E t is the average response delay in the t-th round task set, To set the expected delay, W min 、W max They are the minimum and maximum boundary values allowed by the window respectively;

[0051] S53. After each round of scheduling, based on the historical scheduling status in the status data set and the scheduling variable set, predict the resource demand trend of the unassigned tasks in the task set, and update the scoring factor in the scheduling evaluation model according to the prediction result to revise the scheduling scoring result;

[0052] S54. When there is a resource conflict in the device set and some tasks fail to be effectively allocated, only the successfully allocated part of the current scheduling variable set is retained, and the local scheduling function is called to re-execute the MISAPTS scheduling process on the remaining unallocated tasks, and output the updated scheduling variable set.

[0053] Optionally, the S6 specifically includes:

[0054] S61. Extract task identifier T from the task set i , the corresponding task duration τ i and resource demand weight ρi;

[0055] S62. Get task T from the device collection i The current load ratio η of the corresponding device i , communication delay value δ i , and extract the distribution area identifier γ from the edge node i Indicates the node number to which it belongs;

[0056] S63. Construct a priority scoring function:

[0057]

[0058] Among them, P i For task T i The priority score, μ1, μ2, μ3, μ4, μ5 are weight coefficients, τ i For task T i The duration of i is the resource demand weight, η i is the load ratio of the device corresponding to the task, δ i is the communication delay of the task-associated edge node, γ i is the edge node number to which the task belongs, dist(γ i ) represents the edge node γ i Number of network hops between the system and the main control platform;

[0059] S64. Calculate the priority score P of each task in the task set i , bind the task identifier with the corresponding score and generate a task-score comparison table;

[0060] S65. Arrange the task set in descending order according to the task-score comparison table to generate a priority scheduling queue.

[0061] Optionally, the S7 specifically includes:

[0062] S71. In the system integration architecture, when the number of edge nodes is not less than two, a consensus node set is established. All edge nodes join the consensus node set through the identity registration mechanism and have the scheduling variable set consistency judgment and voting functions;

[0063] S72. Assign a consensus weight to each edge node in the consensus node set, wherein the consensus weight is calculated based on the number of historical scheduling successes, the completeness ratio of the state data set, and the stable frequency of the communication link, and generate a consensus weight vector;

[0064] S73. After the master control platform broadcasts the scheduling variable set to the consensus node set, each edge node performs consistency judgment based on the local state data set and generates a consensus response data packet. The consensus response data packet includes the judgment value and the consensus weight identification information corresponding to the edge node;

[0065] S74. After receiving all consensus response data packets, the main control platform calculates a weighted consistency result based on the consensus weight vector. When the weighted result meets the set threshold, the main control platform confirms that the scheduling variable set is a consistent scheduling result and generates a scheduling confirmation tag.

[0066] S75. When an edge node fails to return consensus response data in three consecutive scheduling cycles, or the judgment value of the current edge node is inconsistent with the statistical result of the main control platform for at least two times, or the control execution unit controlled by the current edge node in the device set fails to report status data in two or more consecutive scheduling cycles, the main control platform deletes the current edge node from the consensus node set and registers the current edge node as invalid in the consensus status record table of the main control platform.

[0067] Optionally, the S8 specifically includes:

[0068] S81. After receiving the consistent scheduling result, the main control platform combines the task identifier and the device identifier in the scheduling variable set to generate a control instruction sequence. Each instruction in the control instruction sequence includes a task identifier, a target device identifier, and execution parameters.

[0069] S82. The main control platform sends the control instruction sequence to the control execution units in the device set through the communication channel. The control execution units execute corresponding tasks according to the received instructions.

[0070] S83, the control execution unit continuously uploads execution status data during the task execution cycle, and the main control platform collects the status data and calculates the execution error value ε i =|O i -R i |, where ε i For task T i The execution error value, O iis the actual output value fed back by the control execution unit, R i For task T in the scheduling variable set i The corresponding expected execution value;

[0071] S84, when the execution error value ε corresponding to any task in the task set i When the set tolerance threshold is exceeded or the control execution unit does not return status data in two consecutive cycles, the main control platform determines that the current scheduling has failed;

[0072] S85. After determining that scheduling has failed, the main control platform extracts a subset of unfinished tasks from the task set and a subset of available devices from the device set, and re-calls the improved MISATPS algorithm based on the current scheduling variable set to perform perturbation rescheduling, generates an updated scheduling variable set, and regenerates the control instruction sequence.

[0073] The beneficial effects of the present invention are:

[0074] The present invention realizes the unified access and data collection of lighting systems, air-conditioning systems, power systems, security systems and access control systems by constructing a three-layer collaborative architecture of system access modules, edge nodes and main control platforms, effectively opening up the status data transmission path between multiple types of heterogeneous systems. With the help of edge nodes, the operating status data and control feedback are pre-processed to generate structured status data sets and task queues, which improves the fusion and processing efficiency of the original data. By establishing a mapping relationship between task sets, equipment sets and scheduling variable sets, introducing task-equipment mapping diagrams and scheduling evaluation models, a multi-objective scheduling optimization system covering response delay, resource energy consumption, load balancing and coordination is constructed.

[0075] At the scheduling algorithm level, the improved MISATPS algorithm proposed in the present invention introduces a sliding scheduling window mechanism and a predictive feedback mechanism, so that the scheduling granularity can be dynamically adjusted. The scheduling process can perceive the task status trend and optimize the variable update logic accordingly, thereby significantly enhancing the scheduling adaptability of the task set in a dynamic environment. At the same time, by continuously collecting feedback status during task execution, a closed-loop relationship between the scheduling variable set and the status data set is constructed, so that the scheduling strategy has real-time correction capabilities. In terms of scheduling stability, the present invention introduces a weighted Byzantine fault-tolerant consensus algorithm, establishes a dynamic consensus mechanism between edge nodes, and allocates consensus weights based on the historical behavior of the nodes, thereby achieving consistent generation of scheduling results at the edge level and effectively avoiding the risk of control conflicts in the device set.

[0076] In addition, during the control execution process, the main control platform converts the consistency scheduling results into a control instruction sequence and sends it to the control execution unit, and continuously monitors the execution status data. Once the execution error exceeds the set threshold or a status feedback interruption occurs, the system can automatically trigger the micro-disturbance rescheduling process based on the current scheduling variable set. While maintaining the stability of the overall scheduling structure, it locally adjusts the abnormal task mapping relationship, thereby improving the system's response flexibility and control robustness under abnormal operating conditions.

[0077] Through the coordinated operation of the above mechanisms, the present invention effectively improves the real-time performance of system integrated control, the intelligence of scheduling, and the self-recovery capability of the system. It is suitable for complex control scenarios with multi-task, multi-node, and multi-objective constraints, and has good engineering practicality and scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0079] Figure 1 This is a flow chart of a real-time monitoring system integrated management and control system proposed by the present invention;

[0080] Figure 2 A flowchart for task-device mapping and scheduling variables construction for a real-time monitoring system integrated management and control system proposed by the present invention;

[0081] Figure 3 This is a flow chart of dynamic scheduling and rescheduling processing of an improved MISAPTS algorithm for a real-time monitoring system integrated management and control system proposed by the present invention. DETAILED DESCRIPTION

[0082] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0083] refer to Figure 1-3 , a real-time monitoring system integrated management and control system, including:

[0084] System access module, used to access lighting, air conditioning, power, security and access control subsystems, configure communication interfaces and control execution units, establish communication connections with the main control platform through edge nodes, and collect operating status data and control feedback;

[0085] The data processing module is used to perform denoising, alignment and caching operations on the collected data by the edge nodes to generate status data sets and task queues;

[0086] Modeling module, used to build task sets and equipment sets, generate task-equipment mapping graphs, and initialize scheduling variable sets;

[0087] The scheduling module is used to set the optimization objective function, build a scheduling evaluation model, and allocate the task set and update the scheduling variable set based on the improved MISAPTS algorithm combined with the sliding scheduling window and prediction feedback mechanism;

[0088] The priority control module is used to generate a priority scheduling queue based on the task urgency, device status and communication delay score;

[0089] The consensus module is used to execute the weighted Byzantine fault-tolerant consensus algorithm when the number of edge nodes is not less than two, and generate consistent scheduling results for the scheduling variable set;

[0090] The instruction execution module is used by the main control platform to generate control instruction sequences and send them to the control execution unit, and receive feedback status data for execution monitoring;

[0091] The rescheduling module is used to trigger the rescheduling process based on the current scheduling variable set under abnormal status or interrupt conditions and regenerate control instructions.

[0092] By establishing a communication connection relationship between the system access module, edge nodes and the main control platform, the present invention realizes the unified access and data collection of the lighting, air conditioning, power, security and access control subsystems, can efficiently collect operating status data and control feedback, and support the unified information foundation of the entire task scheduling process.

[0093] In this embodiment, the modules are connected through the following methods:

[0094] S1. Build a system integration architecture, deploy system access modules, and establish a communication connection with the main control platform through edge nodes;

[0095] S2, collects the operating status data and control feedback of the system access module, and generates a status data set and task queue after processing by the edge node;

[0096] S3. Build a task set and a device set based on the state data set and the task queue, generate a task-device mapping graph, and initialize a scheduling variable set;

[0097] S4. Set the optimization objective function and build a scheduling evaluation model by combining the task-equipment mapping diagram and the scheduling variable set;

[0098] S5. Introduce the improved MISAPTS algorithm, combine the sliding scheduling window with the prediction feedback mechanism, allocate the task set and update the scheduling variable set according to the scheduling evaluation model;

[0099] S6. Build a priority scoring mechanism to adjust the scheduling order based on the urgency of the task set, the status of the device set, and communication delay;

[0100] S7. When the number of edge nodes is not less than two, a weighted Byzantine fault-tolerant consensus algorithm is used to perform consensus calculation on the scheduling variable set to generate a consistent scheduling result.

[0101] S8. Convert the consistency scheduling result into a control instruction, which is sent by the main control platform to the control execution unit in the device set. When an execution exception occurs, the rescheduling process is triggered.

[0102] The present invention constructs a system integration architecture, collects status data, models task sets and equipment sets, establishes a scheduling mechanism and executes control instructions through method steps, forming a complete system control chain from status data set perception to closed-loop execution of scheduling instructions, which significantly improves the coherence and response efficiency of system integration control.

[0103] In this embodiment, the system access module includes a lighting system access unit, an air-conditioning system access unit, a power system access unit, a security system access unit and an access control system access unit. Each access unit is configured with a communication interface and a control execution unit, and establishes a data communication connection with the main control platform through an edge node for collecting operating status data, environmental parameters and control feedback signals.

[0104] The present invention sets the specific structure of the system access module, clearly divides the lighting system access unit, air-conditioning system access unit, power system access unit, security system access unit and access control system access unit, so that the access structure of various subsystems is standardized, the communication path is clear, and the control execution process has good module decoupling.

[0105] In this embodiment, the improved MISAPTS algorithm includes:

[0106] A sliding scheduling window mechanism is introduced to dynamically update the length and stride of the sliding scheduling window based on the execution delay ratio of the task set, the resource utilization rate of the device set, and the adjustment deviation of the scheduling variable set in the previous scheduling cycle;

[0107] Embed a prediction feedback mechanism to predict the status of upcoming tasks in the task set based on the short-term change trend of the status dataset, and embed the prediction results as correction factors in the scheduling evaluation model;

[0108] When the control execution unit fails or there is an abnormal state in the device set, the perturbation rescheduling mechanism is triggered, the scheduling variable set is locally adjusted, and the task set is redistributed only within the corresponding device neighborhood, leaving the rest of the scheduling structure unchanged.

[0109] The present invention refines the construction method of task sets and device sets in task modeling, uses the task-device mapping diagram to complete the constraint expression and adaptation mapping between tasks and resources, and establishes an extensible, computable, and dynamically updateable scheduling structure foundation through scheduling variable set initialization.

[0110] In this embodiment, S3 specifically includes:

[0111] S31, extracting task information based on the status data set and constructing a task set, where each task in the task set includes a task identifier, resource requirements, an estimated duration, and task dependencies;

[0112] S32. Construct a device set based on the device information of the system access module, where each device in the device set includes a device identifier, processing performance, current status, and communication delay parameters;

[0113] S33. Generate a task-device mapping graph based on the adaptation relationship between the task set and the device set. The task-device mapping graph is a bipartite graph structure. The existence of an edge indicates that the task can be executed on the corresponding device.

[0114] S34, initialize the scheduling variable set, the scheduling variable includes the task identifier T i , Equipment Identification D j , scheduling relationship variable x ij , where x ij =1 indicates task T i Assigned to device D j , otherwise x ij =0.

[0115] The present invention sets a multi-objective optimization function that integrates response delay, resource energy consumption, load balancing and collaborative constraints, and constructs a scheduling evaluation model based on the task-device mapping diagram and the scheduling variable set, thereby achieving quantifiable modeling of task scheduling schemes and evaluation support capabilities under multi-dimensional indicator control.

[0116] In this embodiment, the S4 specifically includes:

[0117] S41. Set the scheduling optimization objective function:

[0118]

[0119] Among them, F is the scheduling optimization objective function value, λ1, λ2, λ3, λ4 are weight coefficients, n is the number of tasks in the task set, m is the number of devices in the device set, ω i For task T in the task set i The urgency factor, D i For task T i The response delay, θj For device D in the device collection j The unit energy consumption coefficient, E j For device D j Energy consumption, Var(U j ) represents the device load value U in the device set j The variance of |Δ| is the number of task pairs that meet the coordination constraints in the completed task set, and |Φ| is the total number of task pairs that need to be coordinated in the task set;

[0120] S42, constructing a scheduling evaluation model based on the task-device mapping diagram and the scheduling variable set, the scheduling evaluation model is based on the task identifier T in the task set i , device identifier D in the device set j With the scheduling variable x ij As input, the optimization objective function F is used to represent the scoring result of the current scheduling scheme;

[0121] S43. The scheduling evaluation model is used to evaluate the allocation scheme between the task set and the equipment set during the scheduling process, and serves as a basis for task scheduling decisions.

[0122] The present invention introduces an improved MISAPTS algorithm consisting of a sliding scheduling window mechanism and a prediction feedback mechanism, which realizes dynamic adjustment of scheduling granularity, feedforward correction of state trends and continuous updating of scheduling variable sets during the scheduling process, effectively improving the timeliness, adaptability and reconfigurability of scheduling.

[0123] In this embodiment, the S5 specifically includes:

[0124] S51. Calculate an initial scheduling score for each task in the task set according to the scheduling evaluation model, call the improved MISAPTS algorithm, combine the task-device mapping graph with the scheduling variable set, initialize the first round of task allocation plan and generate the scheduling variable set;

[0125] S52. Introduce a sliding scheduling window mechanism in the scheduling process, set the initial window length and window step, select only a subset of tasks within the window range for allocation in each round of scheduling, and record the scheduling feedback index E. t Then execute window update:

[0126]

[0127] Among them, W t+1 is the length of the next sliding window, W t is the current window length, η is the adjustment coefficient, E t is the average response delay in the t-th round task set, To set the expected delay, W min 、W maxThey are the minimum and maximum boundary values allowed by the window respectively;

[0128] S53. After each round of scheduling, based on the historical scheduling status in the status data set and the scheduling variable set, predict the resource demand trend of the unassigned tasks in the task set, and update the scoring factor in the scheduling evaluation model according to the prediction result to revise the scheduling scoring result;

[0129] S54. When there is a resource conflict in the device set and some tasks fail to be effectively allocated, only the successfully allocated part of the current scheduling variable set is retained, and the local scheduling function is called to re-execute the MISAPTS scheduling process on the remaining unallocated tasks, and output the updated scheduling variable set.

[0130] The present invention builds a priority scoring mechanism, integrates key factors such as task urgency, equipment status and communication delay to generate a task-score comparison table, and establishes a priority scheduling queue based on this, so that the scheduling order has real-time adjustable capabilities and task response level identification capabilities, thereby improving resource allocation accuracy.

[0131] In this embodiment, S6 specifically includes:

[0132] S61. Extract task identifier T from the task set i , the corresponding task duration τ i and resource demand weight ρ i ;

[0133] S62. Get task T from the device collection i The current load ratio η of the corresponding device i , communication delay value δ i , and extract the distribution area identifier γ from the edge node i Indicates the node number to which it belongs;

[0134] S63. Construct a priority scoring function:

[0135]

[0136] Among them, P i For task T i The priority score, μ1, μ2, μ3, μ4, μ5 are weight coefficients, τ i For task T i The duration of i is the resource demand weight, η i is the load ratio of the device corresponding to the task, δ i is the communication delay of the task-associated edge node, γ i is the edge node number to which the task belongs, dist(γ i ) represents the edge node γi Number of network hops between the system and the main control platform;

[0137] S64. Calculate the priority score P of each task in the task set i , bind the task identifier with the corresponding score and generate a task-score comparison table;

[0138] S65. Arrange the task set in descending order according to the task-score comparison table to generate a priority scheduling queue.

[0139] The present invention constructs a consensus node set, adopts a weighted Byzantine fault-tolerant consensus algorithm to perform consistency judgment on the scheduling variable set, and assigns weights according to the historical behavior of edge nodes, thereby achieving consensus generation of edge-level scheduling results and elimination of abnormal nodes, thereby improving scheduling stability and security.

[0140] In this embodiment, the S7 specifically includes:

[0141] S71. In the system integration architecture, when the number of edge nodes is not less than two, a consensus node set is established. All edge nodes join the consensus node set through the identity registration mechanism and have the scheduling variable set consistency judgment and voting functions;

[0142] S72. Assign a consensus weight to each edge node in the consensus node set, wherein the consensus weight is calculated based on the number of historical scheduling successes, the completeness ratio of the state data set, and the stable frequency of the communication link, and generate a consensus weight vector;

[0143] S73. After the master control platform broadcasts the scheduling variable set to the consensus node set, each edge node performs consistency judgment based on the local state data set and generates a consensus response data packet. The consensus response data packet includes the judgment value and the consensus weight identification information corresponding to the edge node;

[0144] S74. After receiving all consensus response data packets, the main control platform calculates a weighted consistency result based on the consensus weight vector. When the weighted result meets the set threshold, the main control platform confirms that the scheduling variable set is a consistent scheduling result and generates a scheduling confirmation tag.

[0145] S75. When an edge node fails to return consensus response data in three consecutive scheduling cycles, or the judgment value of the current edge node is inconsistent with the statistical result of the main control platform for at least two times, or the control execution unit controlled by the current edge node in the device set fails to report status data in two or more consecutive scheduling cycles, the main control platform deletes the current edge node from the consensus node set and registers the current edge node as invalid in the consensus status record table of the main control platform.

[0146] The present invention generates a control instruction sequence from the consistency scheduling result through the main control platform and sends it to the control execution unit, while continuously monitoring the task execution status. When abnormal feedback or state loss is detected, the perturbation rescheduling process is triggered in time to ensure the stability and robustness of the task execution chain.

[0147] In this embodiment, the S8 specifically includes:

[0148] S81. After receiving the consistent scheduling result, the main control platform combines the task identifier and the device identifier in the scheduling variable set to generate a control instruction sequence. Each instruction in the control instruction sequence includes a task identifier, a target device identifier, and execution parameters.

[0149] S82. The main control platform sends the control instruction sequence to the control execution units in the device set through the communication channel. The control execution units execute corresponding tasks according to the received instructions.

[0150] S83, the control execution unit continuously uploads execution status data during the task execution cycle, and the main control platform collects the status data and calculates the execution error value ε i =|O i -R i |, where ε i For task T i The execution error value, O i is the actual output value fed back by the control execution unit, R i For task T in the scheduling variable set i The corresponding expected execution value;

[0151] S84, when the execution error value ε corresponding to any task in the task set i When the set tolerance threshold is exceeded or the control execution unit does not return status data in two consecutive cycles, the main control platform determines that the current scheduling has failed;

[0152] S85. After determining that scheduling has failed, the main control platform extracts a subset of unfinished tasks from the task set and a subset of available devices from the device set, and re-calls the improved MISATPS algorithm based on the current scheduling variable set to perform perturbation rescheduling, generates an updated scheduling variable set, and regenerates the control instruction sequence.

[0153] The present invention uses a micro-disturbance rescheduling mechanism to partially correct the allocation relationship of abnormal tasks while maintaining the basic stability of the original scheduling structure, and quickly completes the generation and issuance of new control instructions in combination with the current scheduling variable set, effectively shortening the fault response time and avoiding full restart of scheduling.

[0154] Example 1:

[0155] To validate the application capabilities of the proposed real-time monitoring system integration and comprehensive management and control system in a city-level integrated building management and control environment, the system was deployed at a municipal operation and support center in Yuexiu District, Guangzhou City, Guangdong Province. This facility comprises five office buildings and three computer room buildings, operating systems including building lighting, variable-frequency air conditioning, emergency power distribution, access control, and video security. The original systems were independently managed by multiple vendors' platforms, presenting typical integration challenges such as inconsistent interfaces, high command response latency, complex device status feedback links, and decentralized exception handling.

[0156] Based on this invention, the project implementation team built a complete system architecture, including deploying a system access module for connecting five types of subsystems, configuring a communication interface to connect to the original system control execution unit, and uniformly collecting operating status data and control feedback signals through edge nodes. The edge nodes perform denoising, caching, and format alignment operations on the collected data, construct a structured status data set, and generate a task queue in real time in combination with the business scheduling flow.

[0157] The system's main control platform constructs task sets and equipment sets through the modeling module, forms a task-equipment mapping diagram, and initializes the scheduling variable set. The scheduling module introduces the improved MISATPS algorithm, realizes dynamic scheduling partitioning of task sets through the sliding scheduling window mechanism, and periodically corrects the scheduling evaluation model based on the short-term trend changes of the status data set, thereby realizing dynamic adaptive adjustment of the multi-objective optimization function. The system can output the scheduling variable set in real time and continuously iterate.

[0158] When the number of edge nodes is not less than two, the consensus module performs consensus judgment on the set of scheduling variables in each round based on the weighted Byzantine fault-tolerant consensus algorithm. The weight calculation comprehensively considers the node task completion rate, communication success rate and scheduling deviation statistics to ensure that the consistent scheduling results generated by the main control platform are stable and resistant to abnormal interference.

[0159] The main control platform then converts the consistency scheduling results into a control instruction sequence, and sends it to the control execution units in each device set through the instruction execution module, while continuously monitoring their status feedback. When there is a significant delay in the feedback information, the value deviation exceeds the threshold, or the cycle is missing, the system immediately starts the rescheduling module. On the basis of retaining the original task allocation structure, it regenerates the scheduling variable set and control instructions only for the abnormal related tasks through local perturbation, thereby realizing rapid recovery and precise correction of the control link.

[0160] The system has been online for three weeks, and the core indicator data of the control system operation are shown in Table 1:

[0161] Table 1 Comparison of operating performance before and after deployment of control system integrated scheduling

[0162]

[0163]

[0164] The data in the table show that the overall response efficiency of the system has increased by more than 60%, the accuracy of status monitoring has increased significantly, the scheduling conflicts between nodes have been basically eliminated, and the delay in abnormal task processing has been shortened by more than five times. The system has achieved practical results in real-time monitoring, unified scheduling, multi-system compatibility and fault-tolerant processing. It is suitable for large-scale operating environments that require real-time comprehensive management, such as urban management platforms, smart buildings, and industrial parks, and has broad engineering application value.

[0165] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A real-time monitoring system integrated management and control system, characterized in that: include: System access module, used to access lighting, air conditioning, power, security and access control subsystems, configure communication interfaces and control execution units, establish communication connections with the main control platform through edge nodes, and collect operating status data and control feedback; The data processing module is used to perform denoising, alignment and caching operations on the collected data by the edge nodes to generate status data sets and task queues; Modeling module, used to build task sets and equipment sets, generate task-equipment mapping graphs, and initialize scheduling variable sets; The scheduling module is used to set the optimization objective function, build a scheduling evaluation model, and allocate the task set and update the scheduling variable set based on the improved MISAPTS algorithm combined with the sliding scheduling window and prediction feedback mechanism; The priority control module is used to generate a priority scheduling queue based on the task urgency, device status and communication delay score; The consensus module is used to execute the weighted Byzantine fault-tolerant consensus algorithm when the number of edge nodes is not less than two, and generate consistent scheduling results for the scheduling variable set; The instruction execution module is used by the main control platform to generate control instruction sequences and send them to the control execution unit, and receive feedback status data for execution monitoring; The rescheduling module is used to trigger the rescheduling process based on the current scheduling variable set under abnormal status or interrupt conditions and regenerate control instructions.

2. A real-time monitoring system integrated management and control system according to claim 1, characterized in that: The modules are implemented as follows: S1. Build a system integration architecture, deploy system access modules, and establish a communication connection with the main control platform through edge nodes; S2, collects the operating status data and control feedback of the system access module, and generates a status data set and task queue after processing by the edge node; S3. Build a task set and a device set based on the state data set and the task queue, generate a task-device mapping graph, and initialize a scheduling variable set; S4. Set the optimization objective function and build a scheduling evaluation model by combining the task-equipment mapping diagram and the scheduling variable set; S5. Introduce the improved MISAPTS algorithm, combine the sliding scheduling window with the prediction feedback mechanism, allocate the task set and update the scheduling variable set according to the scheduling evaluation model; S6. Build a priority scoring mechanism to adjust the scheduling order based on the urgency of the task set, the status of the device set, and communication delay; S7. When the number of edge nodes is not less than two, a weighted Byzantine fault-tolerant consensus algorithm is used to perform consensus calculation on the scheduling variable set to generate a consistent scheduling result. S8. Convert the consistency scheduling result into a control instruction, which is sent by the main control platform to the control execution unit in the device set. When an execution exception occurs, the rescheduling process is triggered.

3. A real-time monitoring system integrated management and control system according to claim 2, characterized in that: The system access module includes a lighting system access unit, an air-conditioning system access unit, a power system access unit, a security system access unit and an access control system access unit. Each access unit is equipped with a communication interface and a control execution unit, and establishes a data communication connection with the main control platform through an edge node to collect operating status data, environmental parameters and control feedback signals.

4. A real-time monitoring system integrated management and control system according to claim 2, characterized in that: The improved MISAPTS algorithm includes: A sliding scheduling window mechanism is introduced to dynamically update the length and stride of the sliding scheduling window based on the execution delay ratio of the task set, the resource utilization rate of the device set, and the adjustment deviation of the scheduling variable set in the previous scheduling cycle; Embed a prediction feedback mechanism to predict the status of upcoming tasks in the task set based on the short-term change trend of the status dataset, and embed the prediction results as correction factors in the scheduling evaluation model; When the control execution unit fails or there is an abnormal state in the device set, the perturbation rescheduling mechanism is triggered, the scheduling variable set is locally adjusted, and the task set is redistributed only within the corresponding device neighborhood, leaving the rest of the scheduling structure unchanged.

5. The system integration and comprehensive management control system for real-time monitoring according to claim 2 is characterized in that: The S3 specifically includes: S31, extracting task information based on the status data set and constructing a task set, where each task in the task set includes a task identifier, resource requirements, an estimated duration, and task dependencies; S32. Construct a device set based on the device information of the system access module, where each device in the device set includes a device identifier, processing performance, current status, and communication delay parameters; S33. Generate a task-device mapping graph based on the adaptation relationship between the task set and the device set. The task-device mapping graph is a bipartite graph structure. The existence of an edge indicates that the task can be executed on the corresponding device. S34, initialize the scheduling variable set, the scheduling variable includes the task identifier T i , Equipment Identification D j , scheduling relationship variable x ij , where x ij =1 indicates task T i Assigned to device D j , otherwise x ij =0.

6. A real-time monitoring system integrated management and control system according to claim 2, characterized in that: The S4 specifically includes: S41. Set the scheduling optimization objective function: Among them, F is the scheduling optimization objective function value, λ1, λ2, λ3, λ4 are weight coefficients, n is the number of tasks in the task set, m is the number of devices in the device set, ω i For task T in the task set i The urgency factor, D i For task T i The response delay, θ j For device D in the device collection j The unit energy consumption coefficient, E j For device D j Energy consumption, Var(U j ) represents the device load value U in the device set j The variance of |Δ| is the number of task pairs that meet the coordination constraints in the completed task set, and |Φ| is the total number of task pairs that need to be coordinated in the task set; S42, constructing a scheduling evaluation model based on the task-device mapping diagram and the scheduling variable set, the scheduling evaluation model is based on the task identifier T in the task set i , device identifier D in the device collection j With the scheduling variable x ij As input, the optimization objective function F is used to represent the scoring result of the current scheduling scheme; S43. The scheduling evaluation model is used to evaluate the allocation scheme between the task set and the equipment set during the scheduling process, and serves as a basis for task scheduling decisions.

7. The system integration and comprehensive management control system for real-time monitoring according to claim 2 is characterized in that: The S5 specifically includes: S51. Calculate an initial scheduling score for each task in the task set according to the scheduling evaluation model, call the improved MISAPTS algorithm, combine the task-device mapping graph with the scheduling variable set, initialize the first round of task allocation plan and generate the scheduling variable set; S52. Introduce a sliding scheduling window mechanism in the scheduling process, set the initial window length and window step, select only a subset of tasks within the window range for allocation in each round of scheduling, and record the scheduling feedback index E. t Then execute window update: Among them, W t+1 is the length of the next sliding window, W t is the current window length, η is the adjustment coefficient, E t is the average response delay in the t-th round task set, To set the expected delay, W min 、W max They are the minimum and maximum boundary values allowed by the window respectively; S53. After each round of scheduling, based on the historical scheduling status in the status data set and the scheduling variable set, predict the resource demand trend of the unassigned tasks in the task set, and update the scoring factor in the scheduling evaluation model according to the prediction result to revise the scheduling scoring result; S54. When there is a resource conflict in the device set and some tasks fail to be effectively allocated, only the successfully allocated part of the current scheduling variable set is retained, and the local scheduling function is called to re-execute the MISAPTS scheduling process on the remaining unallocated tasks, and output the updated scheduling variable set.

8. The system integration and comprehensive management control system for real-time monitoring according to claim 2 is characterized in that: The S6 specifically includes: S61. Extract task identifier T from the task set i , the corresponding task duration τ i and resource demand weight ρ i ; S62. Get task T from the device collection i The current load ratio η of the corresponding device i , communication delay value δ i , and extract the distribution area identifier γ from the edge node i Indicates the node number to which it belongs; S63. Construct a priority scoring function: Among them, P i For task T i The priority score, μ1, μ2, μ3, μ4, μ5 are weight coefficients, τ i For task T i The duration of i is the resource demand weight, η i is the load ratio of the device corresponding to the task, δ i is the communication delay of the task-associated edge node, γ i is the edge node number to which the task belongs, dist(γ i ) represents the edge node γ i Number of network hops between the system and the main control platform; S64. Calculate the priority score P of each task in the task set i , bind the task identifier with the corresponding score and generate a task-score comparison table; S65. Arrange the task set in descending order according to the task-score comparison table to generate a priority scheduling queue.

9. The system integration and comprehensive management control system for real-time monitoring according to claim 2 is characterized in that: The S7 specifically includes: S71. In the system integration architecture, when the number of edge nodes is not less than two, a consensus node set is established. All edge nodes join the consensus node set through the identity registration mechanism and have the scheduling variable set consistency judgment and voting functions; S72. Assign a consensus weight to each edge node in the consensus node set, wherein the consensus weight is calculated based on the number of historical scheduling successes, the completeness ratio of the state data set, and the stable frequency of the communication link, and generate a consensus weight vector; S73. After the master control platform broadcasts the scheduling variable set to the consensus node set, each edge node performs consistency judgment based on the local state data set and generates a consensus response data packet. The consensus response data packet includes the judgment value and the consensus weight identification information corresponding to the edge node; S74. After receiving all consensus response data packets, the main control platform calculates a weighted consistency result based on the consensus weight vector. When the weighted result meets the set threshold, the main control platform confirms that the scheduling variable set is a consistent scheduling result and generates a scheduling confirmation tag. S75. When an edge node fails to return consensus response data in three consecutive scheduling cycles, or the judgment value of the current edge node is inconsistent with the statistical result of the main control platform for at least two times, or the control execution unit controlled by the current edge node in the device set fails to report status data in two or more consecutive scheduling cycles, the main control platform deletes the current edge node from the consensus node set and registers the current edge node as invalid in the consensus status record table of the main control platform.

10. The system integration and comprehensive management control system for real-time monitoring according to claim 2, characterized in that: The S8 specifically includes: S81. After receiving the consistent scheduling result, the main control platform combines the task identifier and the device identifier in the scheduling variable set to generate a control instruction sequence. Each instruction in the control instruction sequence includes a task identifier, a target device identifier, and execution parameters. S82. The main control platform sends the control instruction sequence to the control execution units in the device set through the communication channel. The control execution units execute corresponding tasks according to the received instructions. S83, the control execution unit continuously uploads execution status data during the task execution cycle, and the main control platform collects the status data and calculates the execution error value ε i =|O i -R i |, where ε i For task T i The execution error value, O i is the actual output value fed back by the control execution unit, R i For task T in the scheduling variable set i The corresponding expected execution value; S84, when the execution error value ε corresponding to any task in the task set i When the set tolerance threshold is exceeded or the control execution unit does not return status data in two consecutive cycles, the main control platform determines that the current scheduling has failed; S85. After determining that scheduling has failed, the main control platform extracts a subset of unfinished tasks from the task set and a subset of available devices from the device set, and re-calls the improved MISATPS algorithm based on the current scheduling variable set to perform perturbation rescheduling, generates an updated scheduling variable set, and regenerates the control instruction sequence.

Images