Operation management method and system for smart campus platform

Abstract

The invention relates to the technical field of intelligent campus operation management and control, in particular to an operation management method and system for an intelligent campus platform, and the method comprises the steps: obtaining a personnel flow feature flow, a resource consumption feature flow and a teaching activity feature flow in a campus, and importing a space-time coupled state inference engine, and calculating a comprehensive operation state value of each grid region according to a preset physical constraint rule and a behavior logic rule. A multi-dimensional campus situation tensor is constructed and input into a reverse state inversion network, a key disturbance factor sequence is inverted by comparing current and historical tensor differences, an operation management traceability graph is generated, a trunk causal link is reserved through entropy weight link clipping, and an intervention plan set of key disturbance factors is deduced accordingly. According to the method, time-space fusion quantitative analysis of the multi-source feature flow can be realized, campus operation disturbance core inducements are accurately positioned, a causal traceability structure is simplified, and the matching degree of an intervention plan and actual disturbance inducements is improved.

Description

Technical Field

[0001] This invention relates to the field of smart campus operation and management technology, and in particular to a method and system for the operation and management of a smart campus platform. Background Technology

[0002] Currently, the operation and management of smart campus platforms largely rely on single-dimensional data collection and statistical analysis models. Data on campus personnel flow, resource consumption, and teaching activities are collected and processed independently. Basic data statistics and threshold judgments are used to achieve preliminary monitoring of campus operational status, but a multi-source characteristic stream fusion analysis mechanism has not been established. The analysis processes for various types of data are independent, and a unified spatiotemporal analysis framework has not been formed. While some existing technologies attempt to integrate multiple types of campus operational data, they only reach the level of data aggregation and display, without conducting in-depth calculations based on the correlation between the campus physical space and personnel behavior.

[0003] Such technical solutions can only achieve discrete status monitoring and cannot combine spatiotemporal dimensions to perform unified quantitative calculations of the operational status of the campus grid area. It is difficult to accurately identify the core causes of the campus operational status deviating from the baseline state. At the same time, the causal correlation analysis of multi-source data has a large number of redundant links and cannot automatically filter out the core causal logic. As a result, the operation and management intervention strategies lack accurate source tracing support, and the formulation of intervention plans relies heavily on human experience judgment, resulting in a low degree of matching with the actual operational status of the campus.

[0004] This invention aims to solve the problem of accurately calculating the comprehensive operational status value of a campus grid region under a spatiotemporal coupling framework, and at the same time solve the problems of inverting key disturbance factors and screening the main causal links when the campus operational status deviates from the benchmark. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing an operation and management method and system for a smart campus platform.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an operation and management method for a smart campus platform, comprising: Acquire the characteristic flows of personnel movement, resource consumption, and teaching activities within the campus; The personnel flow feature stream, the resource consumption feature stream, and the teaching activity feature stream are imported into the spatiotemporal coupled state reasoning engine. The spatiotemporal coupled state reasoning engine calculates the comprehensive operating state value of each grid area in the campus at the current moment based on preset physical constraint rules and behavioral logic rules. Based on the comprehensive operating status value, a multi-dimensional campus situation tensor is constructed and input into the inverse state inversion network. The inverse state inversion network analyzes the difference between the current situation tensor and the historical situation tensor to deduce the sequence of key perturbation factors that cause the current state to deviate from the baseline state. Based on the key perturbation factor sequence, an operation management tracing graph containing all potential causal chains is generated. Each node in the operation management tracing graph represents a specific operation event, and each edge represents the causal relationship between events. Perform an entropy-weight-based link pruning operation on the aforementioned operation management traceability map, eliminating redundant causal chains and retaining the main causal chain with the highest entropy weight; Based on the main causal chain, an intervention plan set for each key disturbance factor is derived in reverse. The intervention plan set defines the resource allocation strategy that should be executed under specific triggering conditions.

[0007] As a further aspect of the present invention, the acquisition of the characteristic flow of personnel flow, the characteristic flow of resource consumption, and the characteristic flow of teaching activities within the campus includes: Receive raw operation logs from various business systems within the campus. These raw operation logs include access control records, library borrowing behavior, canteen consumption records, and course attendance data from the academic affairs system. The original operation log is timestamped and normalized, and the personnel flow characteristic stream, resource consumption characteristic stream, and teaching activity characteristic stream are extracted from it; specifically including: The access control records and course attendance data are retrieved and aggregated to calculate the net flow through each pedestrian channel per unit time, forming a set of personnel movement trajectory points, which are then further aggregated into the personnel flow characteristic flow. The canteen's consumption flow and water and electricity meter readings are calculated using a moving average over a unit time window to eliminate instantaneous reading fluctuations. The smoothed data is then converted into instantaneous power and consumption frequency per unit area to form the resource consumption characteristic flow. The course attendance data and classroom occupancy application data of the academic affairs system are correlated and analyzed to extract the real-time timetable load of different teaching buildings and different floors, forming the characteristic flow of the teaching activities; Establish a unified timeline index to map each data record in the personnel flow feature stream, resource consumption feature stream, and teaching activity feature stream to the corresponding time slot of the timeline index.

[0008] As a further aspect of the present invention, the spatiotemporally coupled state reasoning engine calculates the comprehensive operational state value of each grid region within the campus at the current moment based on preset physical constraint rules and behavioral logic rules, including: The entire campus physical space is divided into several equally sized square grid regions, which serve as the basic units for state calculation; For each grid region, the flow characteristics of personnel flow, resource consumption, and teaching activities within the grid region are obtained respectively. Call the preset physical constraint rule library to check whether the personnel flow characteristic flow value exceeds the safe carrying capacity limit of the grid area. If it exceeds, saturate and truncate the personnel flow characteristic flow value. The system calls a pre-set behavioral logic rule library to determine whether the teaching activity feature flow value is within the regular teaching period of the grid area. If it is outside the teaching period, a higher energy consumption coefficient is assigned to the resource consumption feature flow value. The flow values ​​of personnel flow, resource consumption, and teaching activities, after rule correction, are weighted and integrated to output a comprehensive operational status value between zero and one.

[0009] As a further aspect of the present invention, the step of constructing a multi-dimensional campus situation tensor based on the comprehensive operating state value and inputting it into the inverse state inversion network includes: The multi-dimensional campus situation tensor is used to map the distribution of pedestrian density, energy consumption level and teaching load in a three-dimensional spatial coordinate system. The calculated comprehensive operating status values ​​are arranged according to the location coordinates of their corresponding grid areas in the campus geographic space; In a three-dimensional spatial coordinate system composed of longitude, latitude, and floor height, the comprehensive operating status value of each grid region is filled into the corresponding coordinate position as a scalar. In the three-dimensional spatial coordinate system, for each coordinate position filled with the comprehensive operating status value, the original feature flow value that generated the comprehensive operating status value is supplemented and recorded, including the personnel flow feature flow value, resource consumption feature flow value and teaching activity feature flow value of the grid area corresponding to the coordinate position; The entire three-dimensional spatial data volume filled with the comprehensive operating status value and the corresponding original feature flow value is defined as the multi-dimensional campus situation tensor. The constructed multi-dimensional campus situation tensor is transmitted to the input layer of the inverse state inversion network through a preset data interface.

[0010] As a further aspect of the present invention, the inverse state inversion network, by analyzing the difference between the current state tensor and the historical state tensor, deduces the sequence of key perturbation factors that cause the current state to deviate from the baseline state, including: The campus situation tensor of multiple historical moments adjacent to the current moment is retrieved from the time series database, and the mean square error between the current situation tensor and the historical situation tensor is calculated to obtain the state deviation. The state deviation is used as the target output, and the various business data features at the current moment are used as input to train an inversion model of a long short-term memory network structure. During the inversion process, the long short-term memory network structure automatically learns the propagation pattern of state deviation in the time dimension and identifies the top few feature dimensions that contribute the most to the state deviation. The specific business indicators corresponding to the feature dimensions, including a sudden increase in the flow of people at a certain gate or the concentrated activation of air conditioners in a certain dormitory building, are extracted and arranged in chronological order to form the key disturbance factor sequence.

[0011] As a further aspect of the present invention, based on the key perturbation factor sequence, an operation and management tracing map containing all potential causal chains is generated, including: The last factor in the sequence of key perturbation factors is taken as the root node, and the remaining factors are taken as leaf nodes to be associated. Traverse the business knowledge graph of the campus to find the business connection path between the root node and each leaf node. The business connection path consists of a series of intermediate business links. Each found business connection path is converted into a directed causal edge, and the root node, leaf node and intermediate business links are combined into a complete causal chain. All identified causal chains are deduplicated and merged, and duplicate intermediate nodes are removed to finally generate an operational management traceability map that shows the interaction of multiple factors.

[0012] As a further aspect of the present invention, an entropy-weighted link pruning operation is performed on the operation management tracing graph to remove redundant causal chains, including: For each causal chain in the operation management traceability graph, calculate the sum of the information entropy carried by all the nodes it contains to obtain the initial entropy value of the chain; Calculate the average of the initial entropy values ​​of all causal chains, and determine the causal chains whose initial entropy values ​​are lower than the average value as low-information chains; For each causal chain that is determined to be a low-information chain, analyze whether it is completely covered by another high-information chain. If it is completely covered, perform a deletion operation. Reconnect the remaining causal chains after the deletion operation.

[0013] As a further aspect of the present invention, based on the main causal chain, an intervention plan set for each key disturbance factor is derived in reverse, including: From the main causal chain, the upstream business link that directly caused the key disturbance factor was located; Search the campus emergency response plan database to find all standard handling procedures related to the upstream business process, which include adjustment parameters and operational actions for the upstream business process; The standard handling process found is combined with the current specific business parameters to generate specific execution steps for the key disturbance factors. The execution steps constitute the intervention plan set. A priority is assigned to each execution step in the intervention plan set. The priority is determined by the transmission distance of the execution step in the main causal chain, with higher priority for closer steps.

[0014] As a further aspect of the present invention, it also includes: The resource allocation strategy is matched with the current campus resource inventory data to calculate the feasibility index of the strategy implementation, and the intervention plan set is sorted a second time based on the feasibility index. The intervention plan set after secondary sorting is encapsulated into a structured operation scheduling instruction and sent to the underlying equipment control interface to complete a closed-loop operation status adjustment. The resource allocation strategy is matched with the current campus resource inventory data to calculate the feasibility index for implementing the strategy, specifically including: The execution steps of the intervention plan set are analyzed to extract a list of physical resource requirements, which includes the number of personnel, the number of equipment, and the amount of materials consumed. Real-time campus resource inventory data is retrieved from the asset management system interface. The campus resource inventory data includes the current available manpower, equipment and material reserves. Each resource requirement in the requirement list is compared with the corresponding item in the campus resource inventory data to calculate the satisfaction rate, which is the ratio of available resources to required resources. The geometric average of the satisfaction rates of each resource is used as the feasibility index for the implementation of the strategy. When the feasibility index is greater than the preset qualified threshold, the strategy is determined to be executable. The process of encapsulating the secondary sorted set of intervention plans into structured operational scheduling instructions specifically includes: The top three highest-priority execution steps in the intervention plan set are read and used as the core content of this scheduling. An absolute execution timestamp is assigned to each execution step, which is obtained by adding the current system time to the estimated time of the step; The specific details of the execution steps, the target device address, the operation type, and the absolute execution timestamp are encoded according to a predefined communication protocol to generate a standard operation scheduling instruction.

[0015] As a further aspect of the present invention, the present invention also includes an operation and management system for a smart campus platform, the system including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, it implements the steps of the operation and management method for a smart campus platform as described above.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: The spatiotemporally coupled state reasoning engine imports personnel flow feature streams, resource consumption feature streams, and teaching activity feature streams. Based on preset physical constraint rules and behavioral logic rules, it calculates the comprehensive operational state value of each grid area within the campus at the current moment. It achieves unified quantification and transformation of multi-source heterogeneous feature stream data in spatiotemporal dimensions, constructs a spatiotemporal analysis framework that integrates multiple feature streams, eliminates the spatiotemporal dimension disconnect caused by independent analysis of different feature stream data, and makes the quantification results of campus operational status match the campus physical space grid distribution and personnel behavior logic. This enables the quantitative presentation of campus operational status in micro-grid units, improving the accuracy and spatial matching of campus operational status representation.

[0017] A multi-dimensional campus situation tensor input inverse state inversion network is used to inversely deduce the key disturbance factor sequence that causes the current state to deviate from the baseline state by analyzing the difference between the current situation tensor and the historical situation tensor. This completes the targeted extraction of the causes of abnormal campus operation status, filters the influence of non-key disturbance factors, and performs an entropy-weighted link pruning operation on the operation management traceability map generated based on the key disturbance factor sequence. Redundant causal chains in the map are removed, and the main causal links with the highest entropy weights are retained, simplifying the causal relationship structure of the traceability map and forming a clear core causal transmission path. Based on the main causal links, the corresponding intervention plan set for each key disturbance factor is derived in reverse, so that the intervention plan set is directly related to the key disturbance factor. This allows the generation of resource allocation strategies to be based on the core causal logic, improving the correspondence between intervention plans and the actual disturbance causes in campus operation. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the operation and management method of a smart campus platform according to the present invention; Figure 2 A flowchart for calculating the comprehensive operational status value of each grid region; Figure 3 A heat map of the comprehensive operational status of the campus grid area (state reasoning stage); Figure 4 A heat map of the comprehensive operational status of the campus grid area (status calculation stage); Figure 5 This is a spatiotemporal coupling trend diagram of the core campus operation indicators for 24 hours. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] See Figure 1The system continuously acquires real-time characteristic flows of personnel flow, resource consumption, and teaching activities within the campus. These characteristic flows reflect the core dimensions of the campus's dynamic operation. These three types of characteristic flows are imported into a spatiotemporally coupled state reasoning engine. This engine verifies and corrects the characteristic flows based on pre-defined physical constraints and behavioral logic rules that reflect the campus's physical laws and daily behavioral patterns. It then calculates the comprehensive operational state value of each grid area within the campus at the current moment, quantifying the operational load and health of each area. Based on the comprehensive operational state values ​​of all grid areas, a multi-dimensional campus situation tensor is constructed to characterize the overall operational status of the campus. This situation tensor is input into an inverse state inversion network. This network compares the current situation tensor with the situation tensor under historical normal conditions to inversely deduce the sequence of key disturbance factors that cause the current operational state to deviate from the expected or baseline state. Based on the sequence of key disturbance factors and their correlation within the campus business system, an operational management tracing graph containing all potential causal chains is generated. This graph graphically reveals the transmission paths between events. An entropy-weighted link pruning operation is performed on the operational management tracing map. By calculating and comparing the information entropy of each causal chain, causal chains with redundant information or low contribution are eliminated, retaining the main causal chain with the highest entropy weight to focus on the core problem chain. Based on the identified main causal chain, a reverse derivation is performed from the result end to the cause end to generate a set of intervention plans for each key disturbance factor in the chain, defining specific triggering conditions and specific resource allocation strategies, thereby forming a precise management response.

[0022] In one embodiment of the present invention, see [reference] Figure 2 The system continuously receives raw operation logs from the campus access control system, library management system, cafeteria consumption system, and academic affairs system. These raw logs contain timestamp fields with varying formats, requiring timestamp normalization to convert all recorded time information to the same Unix timestamp format within the same time zone. From the normalized log data, the system extracts three core data streams: personnel flow characteristic stream, resource consumption characteristic stream, and teaching activity characteristic stream.

[0023] In practical implementation, generating personnel flow characteristic streams involves processing access control records and course attendance data from the academic affairs system. The system retrieves access control records within a specified time window and aggregates abnormal data from the course attendance data marked as "attended" but without corresponding records generated at the access control near the associated classroom. For each physical pedestrian channel, the difference between the number of people entering and leaving within a unit of time is calculated to obtain the net flow value for that channel. The net flow values ​​of all channels, along with personnel identification and spatial coordinate information, constitute a set of personnel movement trajectory points. The personnel movement trajectory point set is smoothed and serialized and aggregated in the time dimension, ultimately forming a personnel flow characteristic stream indexed by time.

[0024] In some embodiments, a resource consumption feature stream is generated by processing canteen consumption data and water / electricity meter readings. The system reads real-time consumption data from each canteen consumption window and collects water and electricity meter readings installed on various branches of the building. A moving average is calculated for the consumption amount sequence and resource reading sequence within a configurable unit time window. The moving average calculation eliminates reading fluctuations caused by equipment false alarms or instantaneous peaks. The smoothed canteen consumption data is converted into a consumption frequency index per unit area based on the service area of ​​the consumption window. The smoothed water and electricity meter readings are combined with the corresponding monitoring area to convert into an instantaneous power index per unit area. The instantaneous power index per unit area and the consumption frequency index per unit area together constitute the resource consumption feature stream.

[0025] Optionally, generating the teaching activity feature flow requires correlation analysis of course attendance data and classroom occupancy application data from the academic affairs system. The system extracts the planned class times, classroom numbers, and student attendance status for all courses in the academic affairs system, and synchronously reads real-time classroom occupancy application records from the classroom management system. The correlation analysis matches the classroom numbers in the course attendance data with those in the classroom occupancy application records and compares the time information. For each teaching building and each floor within it, the total number of classrooms in the "planned class" or "application occupancy" state at a specific time is counted, and the statistical results are converted into real-time timetable load intensity values. The sequence of real-time timetable load intensity values ​​for all teaching buildings and floors constitutes the teaching activity feature flow.

[0026] In some embodiments, the system establishes a unified timeline index to integrate multi-source feature streams. The unified timeline index uses the system master clock as its source and defines a minimum time granularity based on a time slot length. Each data record in the personnel flow feature stream, resource consumption feature stream, and teaching activity feature stream is mapped to the corresponding time slot in the unified timeline index according to its normalized timestamp. This mapping operation ensures that data from different sources are strictly aligned in the time dimension, forming time-synchronized multi-dimensional feature stream data. In a specific implementation, a spatiotemporally coupled state inference engine calculates the comprehensive operational state value of the grid area. The engine divides the entire campus's digital map into square grids with sides of L meters, with each square grid serving as an independent state calculation unit. For the target grid area, the engine extracts personnel flow feature stream values, resource consumption feature stream values, and teaching activity feature stream values ​​falling within the grid's geographical boundaries from the time-aligned feature stream data.

[0027] In practice, the engine calls a pre-defined physical constraint rule library to check the safety of personnel density in the grid area. The physical constraint rule library defines the safe carrying capacity limit for each type of grid area, which is pre-set based on the area type and size. The engine compares the acquired personnel flow characteristic value, i.e., the real-time number of people, with the safe carrying capacity limit of the target grid area. If the personnel flow characteristic value exceeds the safe carrying capacity limit, the engine performs a saturation truncation on the personnel flow characteristic value, cutting off the excess portion to make the personnel flow characteristic value equal to the safe carrying capacity limit.

[0028] Optionally, the engine invokes a pre-defined behavioral logic rule base to determine the compliance performance consumption of grid activities. The behavioral logic rule base stores the definitions of regular teaching and non-teaching periods for different grid areas. The engine determines whether the teaching activity feature flow value of the target grid area is non-zero and whether the current time is outside the regular teaching period for that grid area. If the teaching activity feature flow value is outside the teaching period, the engine assigns a higher energy consumption coefficient to the resource consumption feature flow value; this energy consumption coefficient is a pre-defined weighting factor greater than 1.

[0029] In practice, the engine performs weighted fusion on the corrected feature flow values. The engine inputs the personnel flow feature flow values ​​(after saturation truncation by the physical constraint rule base), the resource consumption feature flow values ​​(corrected by the behavioral logic rule base coefficients), and the original teaching activity feature flow values ​​into the weighted fusion module. The weighted fusion module calculates the comprehensive operating status value S according to the following formula: ; in: This represents the normalized characteristic flow value of personnel movement. This represents the normalized resource consumption characteristic flow value. This represents the normalized characteristic flow value of the teaching activity. , , These are preset weighting coefficients, and they satisfy... The weighting coefficients reflect the influence weights of different features on the grid's operational state. The calculated comprehensive operational state value S is a scalar value between 0 and 1, and the output is the comprehensive operational state value of the target grid region at the current time.

[0030] In one embodiment of the present invention, a multi-dimensional campus situation tensor is constructed to map the overall operational status of the campus. This multi-dimensional campus situation tensor represents the distribution of pedestrian density, energy consumption levels, and teaching load in a three-dimensional spatial coordinate system composed of longitude, latitude, and floor height dimensions. The construction process begins with spatial organization of the calculated comprehensive operational status values. Based on the latitude and longitude coordinates and floor information of the grid area corresponding to each comprehensive operational status value in the campus geographic space, the system arranges the comprehensive operational status values ​​to their corresponding coordinate positions in the three-dimensional spatial coordinate system. In an example scenario, if a comprehensive operational status value originates from the latitude and longitude coordinates (E116.3°, N39.9°) and is located at a floor height of 3 meters, then this value is filled into the specific coordinate positions identified by the longitude index 116.3, latitude index 39.9, and height index 3 in the three-dimensional spatial coordinate system.

[0031] In practical implementation, to enrich the information hierarchy of the multi-dimensional campus situation tensor, the system supplements the original feature flow values ​​of each coordinate position filled with a comprehensive operational state value in the three-dimensional spatial coordinate system. These supplemented original feature flow values ​​include the personnel flow characteristic flow value, resource consumption characteristic flow value, and teaching activity characteristic flow value of the corresponding grid area. For example, when the coordinates (E116.3°, N39.9°, H3m) are filled with the comprehensive operational state value 0.75, the system simultaneously associates and stores the original data upon which 0.75 is based at that coordinate position: personnel flow characteristic flow value, resource consumption characteristic flow value, and teaching activity characteristic flow value. After completely filling the comprehensive operational state values ​​of all grid areas and their associated original feature flow values ​​into the three-dimensional spatial coordinate system, the resulting three-dimensional spatial data volume, containing spatial coordinates, state scalars, and underlying feature vectors, is defined as the final multi-dimensional campus situation tensor. After construction, the system transmits the multi-dimensional campus situation tensor to the input layer of the inverse state inversion network through a preset data interface.

[0032] In some embodiments, the inverse state inversion network initiates the process of inverting the sequence of key perturbation factors. The inverse state inversion network retrieves campus situation tensors from a time-series database at multiple historical time points adjacent to the current time. These historical campus situation tensors represent the spatial patterns of the system under normal or baseline operating conditions. The network calculates the numerical differences between the current campus situation tensor and the campus situation tensors at each historical time point at the same three-dimensional coordinate position. The state deviation is obtained by calculating the mean square error between the current situation tensor and the historical situation tensors. The formula for calculating the mean square error is: ; in: This represents the calculated state deviation. This represents the total number of all valid coordinate positions in the campus situation tensor. The tensor representing the current state is in the first position. scalar values ​​at each coordinate position The tensor representing a baseline historical state selected from historical moments is at the 1st... Scalar values ​​at each coordinate position. State deviation. It is a scalar that quantifies the overall difference between the current overall operation of the campus and the historical baseline.

[0033] Optionally, the inverse state inversion network uses the obtained state bias as the target output for model training. The network takes various raw business data features collected at the current moment, including but not limited to real-time pedestrian flow at each gate, real-time meter readings, and classroom attendance status, as model input. A reversion model with a long short-term memory (LSM) network structure is trained using historical datasets. The LSM network structure can learn the time-dimensional dependencies and propagation dynamics of state bias. During model training, the LSM network structure automatically captures the propagation patterns and evolution of state bias over continuous time steps through its internal gating mechanism.

[0034] In one embodiment of the present invention, an operation and management tracing graph containing all potential causal chains is generated based on a sequence of key disturbance factors. The system uses the last factor in the key disturbance factor sequence as the starting point, i.e., the root node, for constructing the graph. For example, if the key disturbance factor sequence is ["T1: Passenger flow at the turnstiles in Teaching Building A exceeds the threshold", "T2: Air conditioners in Dormitory Building B are turned on intensively", "Total electricity load in Area C surges"], then "Total electricity load in Area C surges" at the end of the sequence is selected as the root node. The remaining factors in the key disturbance factor sequence that precede "Total electricity load in Area C", namely "Passenger flow at the turnstiles in Teaching Building A exceeds the threshold" and "Air conditioners in Dormitory Building B are turned on intensively", are used as leaf nodes to establish causal relationships with the root node.

[0035] In some embodiments, the system traverses the campus's business knowledge graph to find business connection paths between the root node and each leaf node. The campus's business knowledge graph is a pre-built graph-structured database containing entities within the campus and their business relationships. Starting with the root node "Surge in total electricity load in Area C," the system searches the campus's business knowledge graph for all reachable paths to the leaf nodes "Passenger flow exceeds threshold at the turnstile of Teaching Building A" and "Centralized activation of air conditioners in Dormitory Building B." For example, the search might find a path from the root node "Surge in total electricity load in Area C" to the leaf node "Centralized activation of air conditioners in Dormitory Building B": Area C main distribution box -- power supply ---> Dormitory Building B floor distribution box -- power supply ---> Dormitory Building B 3rd floor air conditioner group. This path consists of a series of intermediate business links such as "Area C main distribution box," "Dormitory Building B floor distribution box," and "Dormitory Building B 3rd floor air conditioner group," forming a candidate business connection path. Each found business connection path depicts a possible causal transmission chain between events.

[0036] In practice, the system converts each identified business connection path into a directed causal edge in the operation management tracing graph. The conversion process involves creating the starting point (root node), the ending point (leaf node), and all intermediate business links along the path as independent nodes in the operation management tracing graph. Then, based on the order of the links in the business connection path, directed edges are drawn between adjacent nodes, pointing from cause to effect. For example, for the business connection path "C Area Main Distribution Box -> B Dormitory Building Floor Distribution Box -> B Dormitory Building 3rd Floor Air Conditioning Group," the system creates nodes for "C Area Main Distribution Box," "B Dormitory Building Floor Distribution Box," and "B Dormitory Building 3rd Floor Air Conditioning Group." A directed edge is drawn between "C Area Main Distribution Box" and "B Dormitory Building Floor Distribution Box," pointing from the former to the latter. Similarly, a directed edge is drawn between "B Dormitory Building Floor Distribution Box" and "B Dormitory Building 3rd Floor Air Conditioning Group." Ultimately, a complete causal chain is formed from the root node "the total power load in area C surges" to the leaf node "the air conditioners in dormitory building B are turned on in a concentrated manner", connected by intermediate nodes and directed edges.

[0037] Optionally, after traversing and transforming the business connection paths between the root node and all leaf nodes, the system obtains multiple sets of nodes and directed edges. The system performs deduplication and merging operations on all identified causal chains. Deduplication involves comparing and merging nodes representing the same entity; for example, nodes identified from different paths, such as "Dormitory Building B Floor Distribution Box," will be merged into one node. Merging involves integrating edges with the same connection relationship to avoid duplicate edges. Through deduplication and merging, an operation management tracing graph showing the interactions of multiple factors, without duplicate nodes or redundant edges, is generated. This operation management tracing graph visualizes the key perturbation propagation network from each leaf node event to the final root node state.

[0038] It is understandable that the generation process of the operation management traceability map can express the determinism of node connections in a formulaic way, and one of the optional connection strength calculations is as follows: ; in: Representing the traceability map of operations management from nodes Pointing to node The connection strength weight of directed edges, Representing the business knowledge graph on campus, from entities To the entity The number of intermediate steps traversed in the shortest business connection path. This value is used to quantify the directness of causal relationships. The smaller the value, the more likely it is to be a node. For nodes The more direct the potential causal effect, the greater its weight in subsequent analysis. The larger the value, the better. However, regardless of the weight, all paths verified through the business knowledge graph will be initially incorporated into the operation management traceability graph, forming a complete network view containing all potential causal chains.

[0039] See Figure 3The heatmap presents a multi-dimensional spatiotemporal distribution of campus operations. Using floors (vertical) and campus areas (horizontal) as a two-dimensional grid base, the comprehensive operational status value (range 0-1) of each grid unit is mapped to a color gradient: blue represents low load (values ​​approaching 0), and red represents high load / high activity (values ​​approaching 1), achieving a visualized and quantitative expression of campus operations. From a data perspective, the comprehensive operational status value of each grid is obtained by weighted fusion and correction of personnel flow characteristics, resource consumption characteristics, and teaching activity characteristics: Regional dimension: Area E maintains the highest status value (0.85-0.95) across all floors, indicating that this area is the core active area of ​​the campus; Area A has the lowest status value (0.10-0.30), indicating the least operational pressure; the status values ​​of Areas B, C, and D increase in a gradient, reflecting the load distribution pattern from the periphery to the core. Floor dimension: The overall status values ​​of floors 1 and 4 are relatively high, while floors 3 and 5 are relatively low, reflecting the differentiated characteristics of different floors in terms of personnel density, energy consumption levels, and teaching load. Heat maps, as two-dimensional projections of the multi-dimensional campus situation tensor, provide an intuitive input visualization carrier for inverse state inversion networks. They can help to quickly locate high-load grid areas and provide spatial decision-making basis for subsequent disturbance factor inversion and operation intervention plan generation.

[0040] In one embodiment of the present invention, an entropy-weighted link pruning operation is performed on the operation management tracing graph to remove redundant causal chains. The system calculates the initial entropy value of each causal chain in the operation management tracing graph. For any causal chain, the system extracts all nodes contained in the causal chain and calculates the information entropy carried by each node. The information entropy is calculated based on the probability distribution of the event corresponding to the node in historical data. The sum of the information entropies of all nodes in a causal chain is the initial entropy value of the causal chain. For example, the initial entropy value of a causal chain containing three nodes—"announcing a large lecture," "increased pedestrian flow at the entrance of the teaching building," and "turning on classroom lights"—is equal to the sum of the information entropies of each of these three nodes.

[0041] In some embodiments, after calculating the initial entropy values ​​for all causal chains, the system calculates the arithmetic mean of these initial entropy values. The system compares the initial entropy value of each causal chain with this arithmetic mean, and classifies causal chains with initial entropy values ​​lower than the arithmetic mean as low-information chains. The classification operation is based on an explicit threshold comparison logic, rather than fuzzy estimation. Referring to Table 1, an exemplary causal chain entropy value calculation and classification result is shown.

[0042] Table 1: Example Table of Initial Entropy Values ​​and Pruning Decisions for Causal Chains

[0043] In practice, for each causal chain identified as a low-information chain, the system further analyzes its coverage relationship with high-information chains. Coverage analysis checks whether the node sequence of the low-information chain completely appears in the node sequence of a causal chain with an initial entropy value higher than the average. For example, the nodes "staggered class dismissal" and "surge in cafeteria traffic" in the low-information chain "staggered class dismissal -> surge in cafeteria traffic" may be completely contained within the high-information chain "school adjusts schedule -> staggered get out of class dismissal -> surge in cafeteria traffic -> accelerated food consumption." If the system confirms that the node sequence of a low-information chain is completely covered by the node sequence of a high-information chain, a deletion operation is performed, removing the low-information chain from the operational management traceability graph. The deletion operation removes redundant causal paths with insufficient information expression.

[0044] Optionally, after analyzing and deleting all low-information chains, the system reconnects the remaining causal chains in the operation management tracing graph. The reconnection process checks for paths that may have been broken due to node deletion and maintains the connectivity of the main causal network by directly connecting existing upstream and downstream nodes or preserving existing complete paths. After link pruning and reconnection, the operation management tracing graph mainly retains causal chains with high initial entropy values ​​and rich information content. These chains constitute the main causal links for subsequent analysis. The formula for calculating the initial entropy value is expressed as: ; in: The initial entropy value represents a single causal chain, and the summation iterates through all the chains contained in that causal chain. 1 node Representing the The probability estimate of the event or state represented by each node within a historical time window, which is obtained from the statistics of historical operation logs.

[0045] Understandably, the process of deriving the intervention plan set based on the main causal link is immediately initiated. The system locates the upstream business link that directly causes the key disturbance factor from the main causal link. For example, if a key disturbance factor is "excessive load on the power grid in Zone C", and the main causal link shows that its direct upstream link is "centralized high-power operation of the air conditioning group in Dormitory Building B", then "centralized high-power operation of the air conditioning group in Dormitory Building B" is located as the direct upstream business link that needs intervention. The system searches the campus emergency plan database to find all standard handling procedures related to "centralized high-power operation of the air conditioning group in Dormitory Building B". The campus emergency plan database stores predefined response scripts for various common campus operation events. The standard handling procedures related to the operation of the air conditioning group may include a set of adjustment parameters and operation actions such as "increasing the temperature setpoint", "switching to energy-saving mode", and "phased start-up and shutdown".

[0046] See Figure 4 In the operational management status calculation phase of the smart campus platform, the campus physical space is divided into a 4x5 grid area. The comprehensive operational status value of each grid area is visualized in the form of a heat map, with the numerical range normalized to the [0,1] interval. The color changes from red to green, corresponding to a high status value. The heat map intuitively reflects the operational load and activity level of different grid areas: High status value areas (green tone, value > 0.6): such as row 2 column 5 (0.79), row 1 column 3 (0.70), row 3 column 4 (0.69), etc., represent areas where the coupling degree between personnel flow, resource consumption, and teaching activities is relatively high, and the area is in a relatively active or high-load operation state. This may correspond to scenarios such as the core area of ​​the teaching building or popular passageways. Low status value areas (red tone, value < 0.3): such as row 3 column 3 (0.20), row 2 column 4 (0.24), row 4 column 3 (0.24), etc., represent areas where the comprehensive operational activity is low, which may be idle areas during non-teaching periods or sparsely populated peripheral areas. Medium-state-value areas (yellow / orange hue, 0.3~0.6): such as row 1, column 1 (0.43), row 2, column 1 (0.53), etc., represent areas where the operational status is at the baseline level, consistent with the normal load characteristics of daily campus operations. Spatially, high-state-value grids exhibit local clustering, reflecting the distribution of hotspot areas in campus operations; low-state-value grids are mostly distributed on the edges or in specific functional areas, reflecting the unevenness of spatial operation. This heatmap provides an intuitive spatial state basis for subsequent construction of the campus situation tensor, inversion of disturbance factors, and formulation of operational intervention strategies. It is a core visualization carrier for state perception and situational analysis in smart campus operation management.

[0047] In one embodiment of the invention, the resource allocation strategy derived from the main causal link is matched with the current campus resource inventory data. The system analyzes each specific execution step in the intervention plan set to extract a list of physical resource requirements. The requirement list clearly lists the types, quantities, and specifications of resources required to execute the intervention action. For example, the requirement list for an execution step "turning on the backup air conditioners in Building A in batches" might include: 2 on-site operators are needed, 3 portable air conditioners with a rated power of 5kW are needed, and 20 meters of cable are needed. The system pulls real-time campus resource inventory data from the asset management system interface. The campus resource inventory data is dynamically updated to show the current number of available human resources on campus, the number of online and idle devices of each type, and the real-time balance of various materials. For example, real-time campus resource inventory data shows that there are currently 5 available security personnel, 4 idle portable air conditioners, and 50 meters of cable remaining.

[0048] In some embodiments, the system compares each resource requirement in the demand list with the corresponding item in the campus resource inventory data to calculate the fulfillment rate. For each resource, the fulfillment rate is calculated as the ratio of the current available quantity of that resource to the required quantity in the demand list. For example, for the demand "2 on-site operators," if there are currently 5 available security personnel, the personnel fulfillment rate is 5 / 2 = 2.5; for the demand "3 portable air conditioners," if there are currently 4 available units, the equipment fulfillment rate is 4 / 3 ≈ 1.33; for the demand "20 meters of cable," if there are currently 50 meters in stock, the material fulfillment rate is 50 / 20 = 2.5. The system performs a geometric mean calculation on all calculated individual resource fulfillment rates, and the result is the feasibility index for strategy implementation. The formula for the geometric mean calculation is expressed as: ; in: This represents the calculated feasibility index for implementing the strategy. This represents the total number of resource items in the requirements list. Representing the The satisfaction rate of the resource This represents a multiplication operation. Based on the example data above, the feasibility index is calculated. The system presets a qualifying threshold; when the calculated feasibility index... If the value exceeds this threshold, the resource allocation strategy is deemed executable. The system then performs a secondary sorting of all execution steps (or their respective strategies) in the intervention plan set based on the calculated feasibility index, with strategies having higher feasibility indices appearing earlier in the sort.

[0049] Optionally, the intervention plan set, after secondary sorting, is encapsulated into a structured operational scheduling instruction. The system reads the top three specific execution steps with the highest feasibility index and priority from the intervention plan set, and uses these three execution steps as the core content of this scheduling. An absolute execution timestamp is assigned to each execution step, obtained by adding the current system time to the estimated execution time of that step. For example, if the current system time is 14:00:00, and the estimated execution time for a step "adjust the lighting brightness in area B to 70%" is 2 minutes, then the absolute execution timestamp for that step is set to 14:02:00. The system encodes the specific content of the execution step, the target device address, the operation type, and the absolute execution timestamp according to a predefined communication protocol encoding standard, thereby generating a machine-readable and executable standard operational scheduling instruction.

[0050] In practice, the coded operational scheduling instructions are sent to underlying terminal devices such as air conditioning controllers, lighting central controllers, and access control managers via the campus IoT platform or dedicated device control interfaces. Upon receiving the instruction, the device controller parses the absolute execution timestamp and operation content, and triggers the corresponding control action at the time specified by the timestamp. For example, if an instruction is sent to the "Central Air Conditioning Unit of the Third Teaching Building," stating "Switch the cooling mode to energy-saving mode at 14:30:00," the air conditioning unit will execute the mode switch precisely at 14:30:00. By issuing a set of structured operational scheduling instructions, the system completes a closed-loop operational status adjustment process, from status analysis, cause tracing, contingency plan generation to resource matching and instruction execution. It can be understood that the generation and issuance of operational scheduling instructions are automated. After calculating the feasibility index and performing a secondary sorting of the intervention plan set, the system automatically triggers the instruction encapsulation and issuance process without manual intervention.

[0051] See Figure 5 In the 24-hour campus core operation indicator trend analysis, the three core indicators—personnel flow, resource consumption, and teaching activity volume—show significant spatiotemporal coupling characteristics and diurnal rhythmic variation patterns. From a temporal perspective, all three indicators form three distinct peak intervals: morning (6-8 am), noon (11 am-1 pm), and evening (4-5 pm), highly consistent with the campus's daily routine. Teaching activity volume peaks at 8 am, 12 pm, and 4 pm, with the peak intervals highly overlapping with peak course times. After 6 pm, it rapidly declines to a low-load level at night, reflecting the strong time constraints of teaching activities. Personnel flow fluctuates synchronously with teaching activity volume, with the peak slightly lagging behind, reflecting the characteristics of pre- and post-class crowd gathering and dispersal. It reaches its highest value at 5 pm, corresponding to the peak crowd flow after school and evening activities. Resource consumption: Consumption increases synchronously with pedestrian flow and teaching activities, peaking at 5 PM. Its curve is smoother, reflecting the lagged response and cumulative effect of resource consumption such as water, electricity, and catering on pedestrian flow and teaching activities. In terms of indicator synergy, the three types of indicators show a strong positive correlation in the peak range, verifying the spatiotemporal coupling characteristics of campus operation: teaching activities drive pedestrian gathering, which in turn triggers a synchronous increase in resource consumption, forming a chain-like transmission logic of "teaching-pedestrian flow-resources." At night (after 8 PM), all three types of indicators remain at low levels, consistent with the campus's low-load operation pattern at night.

[0052] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for the operation and management of a smart campus platform, characterized in that, include: Acquire the characteristic flows of personnel movement, resource consumption, and teaching activities within the campus; The personnel flow feature stream, the resource consumption feature stream, and the teaching activity feature stream are imported into the spatiotemporal coupled state reasoning engine. The spatiotemporal coupled state reasoning engine calculates the comprehensive operating state value of each grid area in the campus at the current moment based on preset physical constraint rules and behavioral logic rules. Based on the comprehensive operating status value, a multi-dimensional campus situation tensor is constructed and input into the inverse state inversion network. The inverse state inversion network analyzes the difference between the current situation tensor and the historical situation tensor to deduce the sequence of key perturbation factors that cause the current state to deviate from the baseline state. Based on the key perturbation factor sequence, an operation management tracing graph containing all potential causal chains is generated. Each node in the operation management tracing graph represents a specific operation event, and each edge represents the causal relationship between events. Perform an entropy-weight-based link pruning operation on the aforementioned operation management traceability map, eliminating redundant causal chains and retaining the main causal chain with the highest entropy weight; Based on the main causal chain, an intervention plan set for each key disturbance factor is derived in reverse. The intervention plan set defines the resource allocation strategy that should be executed under specific triggering conditions.

2. The operation and management method of a smart campus platform as described in claim 1, characterized in that, The acquisition of the characteristic flow of personnel movement, resource consumption, and teaching activities within the campus includes: Receive raw operation logs from various business systems within the campus. These raw operation logs include access control records, library borrowing behavior, canteen consumption records, and course attendance data from the academic affairs system. The original operation log is timestamped and normalized, and the personnel flow characteristic stream, resource consumption characteristic stream, and teaching activity characteristic stream are extracted from it; specifically including: The access control records and course attendance data are retrieved and aggregated to calculate the net flow through each pedestrian channel per unit time, forming a set of personnel movement trajectory points, which are then further aggregated into the personnel flow characteristic flow. The canteen's consumption flow and water and electricity meter readings are calculated using a moving average over a unit time window to eliminate instantaneous reading fluctuations. The smoothed data is then converted into instantaneous power and consumption frequency per unit area to form the resource consumption characteristic flow. The course attendance data and classroom occupancy application data of the academic affairs system are correlated and analyzed to extract the real-time timetable load of different teaching buildings and different floors, forming the characteristic flow of the teaching activities; Establish a unified timeline index to map each data record in the personnel flow feature stream, resource consumption feature stream, and teaching activity feature stream to the corresponding time slot of the timeline index.

3. The operation and management method of a smart campus platform as described in claim 2, characterized in that, The spatiotemporally coupled state reasoning engine calculates the comprehensive operational state value of each grid area within the campus at the current moment based on preset physical constraint rules and behavioral logic rules, including: The entire campus physical space is divided into several equally sized square grid regions, which serve as the basic units for state calculation; For each grid region, the flow characteristics of personnel flow, resource consumption, and teaching activities within the grid region are obtained respectively. Call the preset physical constraint rule library to check whether the personnel flow characteristic flow value exceeds the safe carrying capacity limit of the grid area. If it exceeds, saturate and truncate the personnel flow characteristic flow value. The system calls a pre-set behavioral logic rule library to determine whether the teaching activity feature flow value is within the regular teaching period of the grid area. If it is outside the teaching period, a higher energy consumption coefficient is assigned to the resource consumption feature flow value. The flow values ​​of personnel flow, resource consumption, and teaching activities, after rule correction, are weighted and integrated to output a comprehensive operational status value between zero and one.

4. The operation and management method of a smart campus platform as described in claim 3, characterized in that, The process of constructing a multi-dimensional campus situation tensor based on the comprehensive operational state value and inputting it into the inverse state inversion network includes: The multi-dimensional campus situation tensor is used to map the distribution of pedestrian density, energy consumption level and teaching load in a three-dimensional spatial coordinate system. The calculated comprehensive operating status values ​​are arranged according to the location coordinates of their corresponding grid areas in the campus geographic space; In a three-dimensional spatial coordinate system composed of longitude, latitude, and floor height, the comprehensive operating status value of each grid region is filled into the corresponding coordinate position as a scalar. In the three-dimensional spatial coordinate system, for each coordinate position filled with the comprehensive operating status value, the original feature flow value that generated the comprehensive operating status value is supplemented and recorded, including the personnel flow feature flow value, resource consumption feature flow value and teaching activity feature flow value of the grid area corresponding to the coordinate position; The entire three-dimensional spatial data volume filled with the comprehensive operating status value and the corresponding original feature flow value is defined as the multi-dimensional campus situation tensor. The constructed multi-dimensional campus situation tensor is transmitted to the input layer of the inverse state inversion network through a preset data interface.

5. The operation and management method of a smart campus platform as described in claim 4, characterized in that, The inverse state inversion network analyzes the difference between the current state tensor and the historical state tensor to deduce the sequence of key perturbation factors that cause the current state to deviate from the baseline state, including: The campus situation tensor of multiple historical moments adjacent to the current moment is retrieved from the time series database, and the mean square error between the current situation tensor and the historical situation tensor is calculated to obtain the state deviation. The state deviation is used as the target output, and the various business data features at the current moment are used as input to train an inversion model of a long short-term memory network structure. During the inversion process, the long short-term memory network structure automatically learns the propagation pattern of state deviation in the time dimension and identifies the top few feature dimensions that contribute the most to the state deviation. The specific business indicators corresponding to the feature dimensions, including a sudden increase in the flow of people at a certain gate or the concentrated activation of air conditioners in a certain dormitory building, are extracted and arranged in chronological order to form the key disturbance factor sequence.

6. The operation and management method of a smart campus platform as described in claim 5, characterized in that, Based on the key perturbation factor sequence, an operational management tracing map containing all potential causal chains is generated, including: The last factor in the sequence of key perturbation factors is taken as the root node, and the remaining factors are taken as leaf nodes to be associated. Traverse the business knowledge graph of the campus to find the business connection path between the root node and each leaf node. The business connection path consists of a series of intermediate business links. Each found business connection path is converted into a directed causal edge, and the root node, leaf node and intermediate business links are combined into a complete causal chain. All identified causal chains are deduplicated and merged, and duplicate intermediate nodes are removed to finally generate an operational management traceability map that shows the interaction of multiple factors.

7. The operation and management method of a smart campus platform as described in claim 6, characterized in that, Perform entropy-weighted link pruning on the aforementioned operation management tracing graph to remove redundant causal chains, including: For each causal chain in the operation management traceability graph, calculate the sum of the information entropy carried by all the nodes it contains to obtain the initial entropy value of the chain; Calculate the average of the initial entropy values ​​of all causal chains, and determine the causal chains whose initial entropy values ​​are lower than the average value as low-information chains; For each causal chain that is determined to be a low-information chain, analyze whether it is completely covered by another high-information chain. If it is completely covered, perform a deletion operation. Reconnect the remaining causal chains after the deletion operation.

8. The operation and management method of a smart campus platform as described in claim 7, characterized in that, Based on the aforementioned main causal chain, a set of intervention plans for each key disturbance factor is derived in reverse, including: From the main causal chain, the upstream business link that directly caused the key disturbance factor was located; Search the campus emergency response plan database to find all standard handling procedures related to the upstream business process, which include adjustment parameters and operational actions for the upstream business process; The standard handling process found is combined with the current specific business parameters to generate specific execution steps for the key disturbance factors. The execution steps constitute the intervention plan set. A priority is assigned to each execution step in the intervention plan set. The priority is determined by the transmission distance of the execution step in the main causal chain, with higher priority for closer steps.

9. The operation and management method of a smart campus platform as described in claim 8, characterized in that, Also includes: The resource allocation strategy is matched with the current campus resource inventory data to calculate the feasibility index of the strategy implementation, and the intervention plan set is sorted a second time based on the feasibility index. The intervention plan set after secondary sorting is encapsulated into a structured operation scheduling instruction and sent to the underlying equipment control interface to complete a closed-loop operation status adjustment. The resource allocation strategy is matched with the current campus resource inventory data to calculate the feasibility index for implementing the strategy, specifically including: The execution steps of the intervention plan set are analyzed to extract a list of physical resource requirements, which includes the number of personnel, the number of equipment, and the amount of materials consumed. Real-time campus resource inventory data is retrieved from the asset management system interface. The campus resource inventory data includes the current available manpower, equipment and material reserves. Each resource requirement in the requirement list is compared with the corresponding item in the campus resource inventory data to calculate the satisfaction rate, which is the ratio of available resources to required resources. The geometric average of the satisfaction rates of each resource is used as the feasibility index for the implementation of the strategy. When the feasibility index is greater than the preset qualified threshold, the strategy is determined to be executable. The process of encapsulating the secondary sorted set of intervention plans into structured operational scheduling instructions specifically includes: The top three highest-priority execution steps in the intervention plan set are read and used as the core content of this scheduling. An absolute execution timestamp is assigned to each execution step, which is obtained by adding the current system time to the estimated time of the step; The specific details of the execution steps, the target device address, the operation type, and the absolute execution timestamp are encoded according to a predefined communication protocol to generate a standard operation scheduling instruction.

10. An operation and management system for a smart campus platform, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the operation and management method of a smart campus platform as described in any one of claims 1 to 9.

Images