Intelligent environmental sanitation vehicle collaborative scheduling method and system based on Internet of Things

Abstract

The invention discloses an intelligent environmental sanitation vehicle collaborative scheduling method and system based on the Internet of Things, particularly relates to the technical field of big data analysis and intelligent scheduling, and is used for solving the problems that an existing environmental sanitation vehicle scheduling method is single in optimization target, cannot dynamically collaborate operation timeliness and energy consumption, and is low in efficiency. And the overall energy efficiency is low due to chain reaction caused by local adjustment. Real-time data of environmental sanitation vehicles, operation tasks and energy supply facilities are obtained through the Internet of Things, a resource competition network is dynamically constructed based on the real-time data, statistical significant mutation of topological vulnerability of the resource competition network is recognized, and then linkage influences caused by the mutation are analyzed. A dynamic resource dependence network is constructed based on linkage influence to identify a new key path formed due to betweenness centrality cumulative mutation, and then long-term comprehensive efficiency of different scheduling adjustment strategies on the key path in terms of operation efficiency and energy efficiency is evaluated. And finally, generating and issuing a target collaborative scheduling instruction which enables the overall long-term comprehensive efficiency of the motorcade to be optimal.

Description

Technical Field

[0001] This invention relates to the field of big data analysis and intelligent scheduling technology, and in particular to a method and system for collaborative scheduling of smart sanitation vehicles based on the Internet of Things. Background Technology

[0002] In the field of sanitation operation management, to achieve efficient control of sanitation vehicles, existing vehicle dispatching systems generally adopt monitoring platforms based on geographic information systems and global positioning systems. These platforms can acquire vehicle location information in real time and issue dispatching instructions based on preset fixed routes or simple replanning based on real-time traffic conditions. Meanwhile, with the promotion and application of new energy sanitation vehicles, basic energy constraints such as their driving range are now considered during dispatching. The common practice is to set this as a fixed threshold condition in the path planning algorithm, primarily focusing on completing immediately visible tasks and shortening the distance traveled by each vehicle; its core optimization objective is relatively singular.

[0003] However, when facing large-scale, electrified sanitation vehicle fleets in collaborative dispatch scenarios, it is difficult to effectively coordinate and balance the two key and conflicting objectives of operational timeliness and energy consumption in a dynamically changing environment. Specifically, due to the lack of in-depth analysis and modeling of the complex coupling relationship between real-time vehicle energy consumption dynamics, charging facility status, and operational task requirements, existing methods cannot dynamically evaluate and balance global time efficiency and energy efficiency during dispatch decisions. This leads to the pre-planned so-called optimal dispatch scheme easily failing during real-time execution, which may not only cause operational delays or energy waste, but may also trigger chain reactions due to emergency adjustments to local tasks, causing the overall energy dispatch strategy of the fleet to fall into an inefficient or even infeasible state, thus failing to achieve optimal comprehensive energy efficiency for long-term stable operation of the fleet. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a method and system for collaborative scheduling of smart sanitation vehicles based on the Internet of Things.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: The IoT-based smart sanitation vehicle collaborative scheduling method includes: S1. Obtain real-time vehicle status data of sanitation vehicles, real-time task demand data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. S2. Based on real-time vehicle status data, real-time task requirement data, and real-time energy status data, identify resource competition networks and assess whether their topological vulnerability index change rate has undergone statistically significant abrupt changes. S3. When a statistically significant abrupt change occurs, analyze the resulting cascading effects on the subsequent tasks of other vehicles in the convoy; S4. Construct a dynamic resource dependency network based on cascading effects, and identify consecutive task nodes that have undergone betweenness centrality cumulative mutations due to cascading effects, and determine the task chain formed by them as the new critical path. S5. Based on the new critical path, evaluate the long-term comprehensive effectiveness of each scheduling adjustment strategy by weighing the combined impact of different scheduling adjustment strategies on the operational efficiency and energy efficiency of the new critical path. S6. Generate and issue target collaborative scheduling instructions based on long-term comprehensive performance to optimize the overall long-term comprehensive performance of the fleet.

[0006] Furthermore, S1 includes: The real-time location information, real-time driving speed and real-time power battery state of charge data of sanitation vehicles are obtained as real-time vehicle status data. Obtain the task location coordinates, planned task time window, and estimated task load data of the tasks to be executed as real-time task requirement data; The location coordinates, real-time occupancy status, and real-time available output power data of energy supply facilities are obtained as real-time energy status data.

[0007] Furthermore, S2 includes: Taking sanitation vehicles and tasks to be performed as the first type of nodes, and energy supply facilities and high-conflict road sections as the second type of nodes, a weighted connection edge reflecting the demand of vehicles and tasks for the second type of nodes is established based on the task location coordinates and planned operation time window in the real-time task demand data, and the real-time location information in the real-time vehicle status data, thereby constructing a resource competition network. The spectral radius of the resource contention network over a continuous time window is calculated as a topological vulnerability index, and the rate of change of the topological vulnerability index over time is obtained. The cumulative sum control chart algorithm is used to monitor the rate of change of topological vulnerability indicators over time to determine whether there are statistically significant abrupt changes in the rate of change of topological vulnerability indicators.

[0008] Furthermore, the spectral radius of the resource competition network over a continuous time window is calculated as a topological vulnerability index. This is achieved in the following way: the constructed resource competition network is represented as a weighted adjacency matrix, where the element values ​​are determined by the weights of the corresponding connecting edges; for each continuous time window, all eigenvalues ​​of the weighted adjacency matrix are calculated and their absolute values ​​are taken. The eigenvalue with the largest absolute value is the spectral radius of the resource competition network for the corresponding time window, which is then used as a topological vulnerability index.

[0009] Furthermore, S3 includes: When a statistically significant mutation occurs, based on the resource competition network that has been identified as having a statistically significant mutation, energy supply facilities whose demand exceeds the real-time available output power due to changes in the weight of the connecting edges in the resource competition network are identified as key conflict resources. Determining all planned job time windows depends on pending job tasks that use critical conflicting resources; For each pending task that depends on critical conflict resources, the additional energy replenishment requirement and the offset of the planned operation time window are calculated based on the real-time power battery state of charge data in the real-time vehicle status data and the estimated operation load data in the real-time task demand data. The additional energy replenishment requirements and planned operation time window offsets calculated for all critical conflict resources are defined as cascading effects.

[0010] Furthermore, S4 includes: A dynamic resource dependency network is constructed using the tasks to be executed as nodes, as well as the dependencies between tasks caused by sharing key conflicting resources, overlapping planned operation time window offsets, or conflicting additional energy replenishment needs. Calculate the betweenness centrality of each task node in the dynamic resource dependency network before and after the change in inter-task dependency caused by cascading effects, and obtain its sequence of betweenness centrality change values. The Mann-Kendall trend test algorithm was used to analyze the sequence of betweenness centrality changes for each task node, and consecutive task nodes with statistically significant upward trends in betweenness centrality were identified. The identified consecutive task nodes are connected sequentially according to their corresponding planned operation time windows to form a new critical path.

[0011] Furthermore, the betweenness centrality of each task node in the dynamic resource dependency network before and after the change in inter-task dependencies caused by cascading effects is calculated in the following way: Based on the topology of the constructed dynamic resource dependency network, the shortest path between each pair of task nodes in the network is calculated in both the network states before and after the cascading effects occur; for each task node, the number of shortest paths passing through the corresponding task node in each state is counted, and the ratio of the number of shortest paths to the total number of shortest paths of all task node pairs in the network in the corresponding state is defined as the betweenness centrality of the corresponding task node in this state.

[0012] Furthermore, S5 includes: For the new critical path, generate at least a variety of scheduling adjustment strategies, including adjusting the task execution order, adjusting the allocation of critical conflict resources, and inserting preset charging periods for tasks; Based on real-time vehicle status data, real-time task demand data, and additional energy replenishment demand and planned operation time window offset in the chain reaction, the cumulative operation time and cumulative energy consumption required to complete all tasks in the new critical path after each scheduling adjustment strategy is simulated. For each scheduling adjustment strategy, based on the simulated cumulative operation time and cumulative energy consumption, combined with the preset operation time weight coefficient and energy consumption weight coefficient, a strategy evaluation value representing the long-term comprehensive effectiveness is calculated.

[0013] Furthermore, S6 includes: Compare the policy evaluation values ​​corresponding to all scheduling adjustment policies, and determine the scheduling adjustment policy with the best policy evaluation value as the target scheduling adjustment policy; Based on the target scheduling adjustment strategy, which includes operations such as adjusting the task execution order, adjusting the allocation of key conflicting resources, or inserting preset charging periods, a target collaborative scheduling instruction containing specific execution time, target resources, and action parameters is generated. The target coordinated dispatch instructions are issued to the corresponding sanitation vehicles and energy supply facilities.

[0014] On the other hand, the present invention provides an IoT-based intelligent sanitation vehicle collaborative dispatch system, comprising: The data acquisition module is used to acquire real-time vehicle status data of sanitation vehicles, real-time task requirement data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. The mutation identification module is used to identify resource competition networks and assess whether statistically significant mutations have occurred in the rate of change of their topological vulnerability indicators based on real-time vehicle status data, real-time task requirement data, and real-time energy status data. The cascading analysis module is used to analyze the cascading effects on subsequent tasks of other vehicles in the fleet when a statistically significant mutation occurs. The path determination module is used to construct a dynamic resource dependency network based on cascading effects, identify consecutive task nodes that have accumulated mutations in betweenness centrality due to cascading effects, and determine the task chain formed by them as a new critical path. The performance evaluation module is used to evaluate the long-term overall performance of each scheduling adjustment strategy based on the new critical path by weighing the combined impact of different scheduling adjustment strategies on the operational efficiency and energy efficiency of the new critical path. The instruction issuance module is used to generate and issue target collaborative scheduling instructions that optimize the overall long-term comprehensive efficiency of the fleet based on long-term comprehensive performance.

[0015] The beneficial effects of this invention are: 1. By constructing a closed-loop collaborative scheduling mechanism from real-time situational awareness to forward-looking decision-making, the shortcomings of existing technologies, such as single optimization objectives and lack of dynamic coordination, are effectively overcome. Through IoT data fusion, two models that reveal the intrinsic relationship of the system—resource competition network and dynamic resource dependence network—are dynamically constructed and analyzed. This allows for a deep characterization of the complex coupling relationship between vehicles, tasks, and facilities as time and state change. By identifying statistically significant mutations and chain effects of network topology vulnerabilities, a shift from passive response to proactive early warning is achieved. This enables the scheduling system to perceive systemic risks in advance. Furthermore, by evaluating the comprehensive impact of different adjustment strategies on the operational efficiency and energy efficiency on new critical paths, dynamic and quantitative trade-offs between conflicting timeliness and energy consumption targets are realized in sanitation vehicle scheduling. This ensures optimal comprehensive energy efficiency at the global level from the source of decision-making, rather than simply optimizing a single path or immediate cost.

[0016] 2. By analyzing the chain reaction and dynamically identifying and reconstructing the critical path, a transmission analysis model between local disturbances and global effectiveness was established. This fundamentally avoids the problem of overall scheduling strategy disorder caused by local emergency adjustments in existing methods. The scheduling decision is based on the analysis of deep network characteristics such as the cumulative mutation of task chain betweenness centrality. This allows the generated target-coordinated scheduling instructions to not only target currently visible conflicts, but also to proactively stabilize the task sequence that has the greatest impact on the overall fleet operating efficiency, thereby suppressing the spread of risks. This enables the scheduling system of the large-scale electric sanitation vehicle fleet to have the resilience and adaptability to cope with dynamic environments. It can continuously output robust scheduling strategies based on long-term comprehensive effectiveness, ensuring that the fleet maintains a high efficiency state that combines high operation completion rate and low energy consumption cost in complex operating environments. Attached Figure Description

[0017] Figure 1 This is a flowchart of the IoT-based smart sanitation vehicle collaborative scheduling method of the present invention; Figure 2 This is a schematic diagram of the structure of the IoT-based smart sanitation vehicle collaborative dispatching system of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Figure 1 This invention presents a smart sanitation vehicle collaborative scheduling method based on the Internet of Things, comprising: S1. Obtain real-time vehicle status data of sanitation vehicles, real-time task demand data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. S2. Based on real-time vehicle status data, real-time task requirement data, and real-time energy status data, identify resource competition networks and assess whether their topological vulnerability index change rate has undergone statistically significant abrupt changes. S3. When a statistically significant abrupt change occurs, analyze the resulting cascading effects on the subsequent tasks of other vehicles in the convoy; S4. Construct a dynamic resource dependency network based on cascading effects, and identify consecutive task nodes that have undergone betweenness centrality cumulative mutations due to cascading effects, and determine the task chain formed by them as the new critical path. S5. Based on the new critical path, evaluate the long-term comprehensive effectiveness of each scheduling adjustment strategy by weighing the combined impact of different scheduling adjustment strategies on the operational efficiency and energy efficiency of the new critical path. S6. Generate and issue target collaborative scheduling instructions based on long-term comprehensive performance to optimize the overall long-term comprehensive performance of the fleet.

[0020] S1. Obtain real-time vehicle status data of sanitation vehicles, real-time task demand data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. Specifically, this is implemented as follows: First, real-time vehicle status data of sanitation vehicles is acquired. This process is achieved through an onboard IoT terminal installed on each sanitation vehicle. This terminal integrates a GPS module, a vehicle controller LAN bus interface, and a battery management system communication interface. Real-time location information is collected and transmitted by the GPS module at a preset frequency. The data format includes coordinate data containing longitude, latitude, and a timestamp. The preset frequency is, for example, once per second or once every 5 seconds, with the specific value configured according to the communication network load and positioning accuracy requirements. Real-time driving speed is read from the vehicle speed sensor through the vehicle controller LAN bus interface and transmitted in kilometers per hour. The data reading cycle is, for example, consistent with the standard message cycle of the vehicle controller LAN bus. Real-time battery state-of-charge (SOC) data is acquired through the battery management system communication interface. The battery management system periodically calculates the percentage of the battery's remaining charge relative to its total capacity. This percentage is the real-time SOC data and is transmitted to the dispatch center's data receiving server via the onboard IoT terminal. The calculation cycle is, for example, once per second. All real-time vehicle status data includes a unique vehicle identifier and a data collection timestamp for subsequent data association and processing. If no data is received from a vehicle within a preset time window, its status is marked as communication interrupted, and the corresponding data retransmission or exception handling process is triggered.

[0021] Secondly, real-time task requirement data for pending tasks is obtained. This data comes from the database of the sanitation work order management system. The dispatch center periodically retrieves the latest work order information from this database via an application programming interface (API), for example, once per minute. Task location coordinates are extracted from the work order information and are typically represented in Geographic Information System (GIS) coordinates, such as latitude and longitude, to identify the specific geographical location where the task needs to be performed. The planned task time window also comes from the work order information. It defines the allowed time range for the task, such as a specific start and end time. These two time points together form a closed time window. The time window is set manually by the task dispatcher based on the management regulations and actual needs of the work area, or automatically generated by the system based on historical data. Estimated workload data is a predefined parameter in the work order that characterizes the amount of work required to complete the task. For example, for sweeping operations, this data could be the product of the length of the road to be swept and the area at standard width, in square meters. The road length and width information is extracted from road layer data in a geographic information system database. For garbage collection operations, this data could be the estimated volume or weight of the garbage bins to be collected, in cubic meters or tons. The volume or weight information is estimated based on the garbage bin's model and specifications and historical collection frequency. Real-time task requirement data is also bound to a unique task identifier and is dynamically updated as work orders are created, modified, or assigned. The dispatch center ensures that it obtains the latest version of the task requirement data by comparing the data version number or timestamp.

[0022] Finally, real-time energy status data of energy supply facilities is obtained. Energy supply facilities mainly refer to charging piles or charging stations that provide charging services for electric sanitation vehicles. This real-time status data is collected and reported through an IoT communication module connected to the charging pile controller. The reporting mechanism can adopt a timed reporting mode or an immediate reporting mode when the status changes. The facility location coordinates are fixed geographical coordinates determined during the deployment of the charging facility, such as the latitude and longitude of its location. This data is pre-stored in the facility information database and entered into the system during facility registration. Real-time occupancy status data indicates whether the charging pile is currently being used by a vehicle. This status is determined by the charging pile controller by detecting the connection status of its charging interface and the status of the charging relay. For example, when the charging gun is correctly connected and the charging process has started, the status is "occupied"; otherwise, it is "idle." This status information is pushed to the dispatch center in real time. Real-time available output power data reflects the maximum charging power that a charging pile can safely and stably provide at the current moment. This data is not a constant value and may change dynamically due to local grid load, multi-pile collaborative scheduling strategies within the charging station, or the facility's own derating operation strategy. Specifically, the charging pile controller calculates the maximum allowable output power value at the current moment based on the input voltage and current monitored by its internal electrical measurement unit, the rated power limit indicated on the facility's nameplate, and real-time power allocation instructions from the higher-level energy management system. This value is then reported in kilowatts. The limit in the power allocation instruction may be a percentage dynamically adjusted according to the total grid load, for example, limiting the available power to 80% of the rated power. Real-time energy status data for all energy supply facilities is associated with a unique facility identifier and continuously updated to the dispatch center's backend database, forming a real-time dynamic perception of energy supply capacity. Through three parallel data acquisition processes, the dispatch system obtains the multi-dimensional, real-time basic data set necessary for subsequent collaborative scheduling analysis. Any anomaly or interruption in any data stream is recorded and alarmed by the monitoring system, ensuring the reliability of the data acquisition process.

[0023] S2. Based on real-time vehicle status data, real-time task requirement data, and real-time energy status data, identify resource contention networks and assess whether their topological vulnerability index change rate has undergone statistically significant abrupt changes. Specifically, this is implemented as follows: First, a resource competition network is constructed. Each sanitation vehicle and each task to be performed is considered a first-type node. Each energy supply facility is considered part of a second-type node. Simultaneously, high-conflict road sections are identified as another part of the second-type nodes. High-conflict road sections are defined as sections where, historically, the average speed has consistently fallen below a preset speed threshold due to multiple sanitation vehicles operating simultaneously or passing through. The preset speed threshold, for example, is 15 kilometers per hour, based on the typical minimum safe driving speed for urban road sanitation operations. The identification method involves dividing the road into grid segments based on real-time location and speed data from historical real-time vehicle status data. The historical average speed of each segment during peak operating hours is calculated, and segments with historical average speeds below the preset speed threshold are marked as high-conflict road sections, with their geographical coordinates taken as the coordinates of the center point of that segment. Based on the task location coordinates and planned operation time window in the real-time task requirement data, a weighted connection edge is established between the task to be executed and the second type of node. Specifically, for a task to be executed, the Euclidean distance between its task location coordinates and the coordinates of each second type of node is calculated. If the distance is less than the first and third association distance thresholds, for example, 500 meters, and the current system time is within the planned operation time window of the task, then a connection edge is established between the task node and the second type of node. The first and third association distance thresholds are set based on the maximum detour radius that sanitation vehicles are usually allowed to detour when they are working or charging. Based on the real-time location information in the real-time vehicle status data, weighted connection edges are established between sanitation vehicles and second-type nodes. Specifically, for a sanitation vehicle, the Euclidean distance between its real-time location information and the coordinates of each second-type node is calculated. If this distance is less than the second and third association distance thresholds, such as 300 meters, a connection edge is established between the vehicle node and the second-type node. The second and third association distance thresholds are less than the first and third association distance thresholds, and their setting is based on the fact that the perception and competitive response distance of a vehicle already in motion to nearby resources is usually shorter than the static association distance of the task planning. The weight values ​​of the connection edges are dynamically set according to both distance and time urgency. For example, the weight value is inversely proportional to the distance value and inversely proportional to the remaining time length of the planned operation time window.

[0024] Secondly, the spectral radius of the resource competition network over continuous time windows is calculated as a topological vulnerability index, and the rate of change sequence of the topological vulnerability index over time is obtained. The constructed mathematical model of the resource competition network is represented as a weighted adjacency matrix. This weighted adjacency matrix is ​​a square matrix, with its rows and columns corresponding to all nodes in the network. The value of an element in the weighted adjacency matrix is ​​determined by whether there is a connecting edge between the corresponding row node and column node and the current weight of that connecting edge; if there is a connecting edge between two nodes, the value of the element is the current weight of that connecting edge; if there is no connecting edge, the value of the element is zero. For each continuous time window, with a length of, for example, 30 minutes, an instantaneous topological snapshot of the resource competition network is extracted at the end of each time window, forming the weighted adjacency matrix corresponding to that moment. For each weighted adjacency matrix, all its eigenvalues ​​are calculated. The calculation of eigenvalues ​​can be implemented using numerical methods, such as the QR iteration method. After calculating all eigenvalues, the absolute value of each eigenvalue is taken, and then the maximum value is found among all absolute values. This maximum absolute value is the spectral radius of the resource contention network under that time window. The spectral radii calculated for each time window are arranged in chronological order to form a sequence of topological vulnerability indicators. Based on this sequence, its rate of change over time is calculated. The rate of change is calculated by the difference between adjacent time points: the spectral radius value of the next time window is subtracted from the spectral radius value of the previous time window, and then divided by the time interval between the two time windows, thus obtaining the sequence of topological vulnerability indicator rate of change.

[0025] Finally, the cumulative sum control chart algorithm is used to monitor the change rate sequence of topological vulnerability indicators to determine whether statistically significant abrupt changes have occurred. First, the reference values ​​for the cumulative sum control chart algorithm, the target mean and process standard deviation, need to be set. The target mean is set as the long-term average of the topological vulnerability indicator change rate sequence under steady-state conditions, obtained by collecting the change rate data of topological vulnerability indicators over a historical stable operating period and calculating its arithmetic mean. The process standard deviation is also calculated based on this historical data set and is used to measure the normal fluctuation range of the change rate. The decision interval control limits for the cumulative sum control chart algorithm are set, consisting of the target mean plus or minus K times the process standard deviation, where K is a multiple of the control limit, for example, K = 3. Then, the real-time calculated change rate sequence of topological vulnerability indicators is monitored online. For each new data point in the sequence, its difference from the target mean is calculated, and this difference is added to the cumulative sum of the previous data point to obtain a new cumulative sum statistic. The cumulative sum statistic is continuously monitored. If its value is greater than the upper limit of the decision interval control limit or less than its lower limit, a statistically significant abrupt change in the rate of change of the topological vulnerability index is determined. The setting of the decision interval control limit, i.e., the selection of the K value, needs to be adjusted according to the requirements for the false alarm probability. The false alarm probability is usually set to a small value, such as 0.3%. Simultaneously, the algorithm includes a reset mechanism; that is, when a sudden change is determined, the cumulative sum statistic is reset to zero so that monitoring can resume.

[0026] S3. When a statistically significant abrupt change occurs, analyze the resulting cascading impact on the subsequent tasks of other vehicles in the convoy. Specifically, this is implemented as follows: First, upon receiving a statistically significant mutation judgment signal from the topology vulnerability index change rate monitoring stage, the critical conflict resource is identified based on the resource competition network where statistically significant mutations have already been identified. Specifically, this identification method involves traversing all energy supply facility nodes in the resource competition network. For each energy supply facility node, the weight values ​​of all weighted edges connected to it at the current moment are aggregated, and a sum is obtained. This sum represents the total occupancy demand of all vehicles and tasks for that energy supply facility node at the current moment. Simultaneously, the real-time available output power data corresponding to that energy supply facility node is obtained from real-time energy status data. The calculated total occupancy demand is compared with the real-time available output power. If the total occupancy demand is greater than the real-time available output power, the service capacity of that energy supply facility node is determined to be insufficient to meet the current competition demand, and it is marked as a critical conflict resource. The comparison between the total occupancy demand and the real-time available output power is the specific implementation of determining whether changes in the edge weights in the resource competition network have caused the occupancy demand to exceed the real-time available output power. Exceeding the demand means that the demand exceeds the supply; this is a clear numerical comparison relationship.

[0027] Secondly, all pending tasks whose planned operation time windows depend on the use of critical conflict resources are identified. This step is based on filtering the planned operation time window information in real-time task demand data and the established connection edges between task nodes and critical conflict resource nodes. For a pending operation task, if there is a connection edge between the task node and a critical conflict resource node in the resource contention network, and the current system time or a near future time point is still within the planned operation time window, then the planned operation time window of the pending operation task is considered to depend on the use of the critical conflict resource. Here, "depends on use" means that the task originally planned to use the critical conflict resource within the planned operation time window to complete the operation or replenish energy. All pending operations tasks that meet the above conditions are identified, forming a set of affected pending operations tasks.

[0028] Next, for each identified task dependent on critical conflict resources, the additional energy replenishment requirement and planned task time window offset due to the inability to use critical conflict resources as originally planned are calculated. The specific method for calculating the additional energy replenishment requirement is as follows: Based on real-time vehicle status data, the real-time battery state-of-charge data reported by the sanitation vehicle scheduled to perform the task is obtained. This data is a percentage value, such as 60%. Based on real-time task requirement data, the estimated workload data for the task is obtained. This data represents the estimated energy required to complete the task, in kilowatt-hours (kWh). For example, the estimated workload data for this task is 5 kWh. First, based on the estimated workload data and the sanitation vehicle's unit work energy efficiency parameter, the electrical energy consumed to complete the task is calculated, in kilowatt-hours. The unit work energy efficiency parameter is the electrical energy consumed by the vehicle per unit of work completed. For example, this parameter is 1.5 kWh consumed per 1000 square meters cleaned. This parameter can be obtained through statistical analysis of the vehicle's historical work data. Then, based on the real-time state-of-charge (SOC) data of the power battery and the total capacity parameter of the vehicle's power battery, the available energy value stored in the vehicle's current battery is calculated, in kilowatt-hours (kWh). For example, if the total battery capacity is 100 kWh and the real-time SOC data is 60%, then the current available energy value is 60 kWh. Next, it is determined whether the energy value required to complete the task is greater than the available energy value of the vehicle's current battery. If it is greater, the difference is calculated as the theoretical energy gap. Since the originally planned critical conflict resource is unavailable, the vehicle needs to travel to other energy supply facilities for replenishment. This travel process itself will also consume a certain amount of driving energy. This driving energy consumption is calculated based on the estimated travel distance based on the location coordinates of the alternative energy supply facilities and the vehicle's real-time location information, as well as the vehicle's unit distance driving energy consumption parameter; the unit distance driving energy consumption parameter is, for example, 0.2 kWh per kilometer. Finally, the above theoretical energy gap is added to the additional driving energy consumption generated by traveling to alternative facilities to obtain the final additional energy replenishment requirement value for the task, in kilowatt-hours. The specific method for calculating the planned operation time window offset is as follows. The original planned operation time window has a starting time point. When critical conflict resources become unavailable, the task needs to adjust its execution plan. The new plan may involve waiting for resources to become available or traveling to an alternative facility. The new planned starting time point is determined based on resource availability forecasts and travel time estimates. For example, if waiting is chosen, the new starting time point is predicted to be the time when the critical conflict resources are expected to become available based on their current occupancy. This predicted time can be obtained by analyzing the historical patterns of real-time occupancy status of charging stations. If traveling to an alternative facility is chosen, the new starting time point is the current time plus the travel time required to travel to the alternative facility, calculated based on the estimated travel distance and the vehicle's average speed.The time window offset for a planned task is defined as the time difference between the start time of the new plan and the start time of the original plan. The unit can be minutes. This offset represents the length of time that the task execution is delayed.

[0029] Finally, the additional energy replenishment requirements and planned operation time window offsets for all affected pending tasks, calculated from all key conflict resources, will be summarized and recorded. This summary information set, containing all specific values, is defined as the cascading effect. This cascading effect quantitatively characterizes the specific impact and scope on the subsequent tasks of other vehicles in the fleet in terms of energy replenishment and time planning caused by a statistically significant mutation in the resource competition network, providing a clear quantitative input basis for subsequent scheduling adjustments.

[0030] S4. Construct a dynamic resource dependency network based on cascading effects, and identify consecutive task nodes that experience cumulative mutations in betweenness centrality due to cascading effects. Determine the task chain formed by these nodes as the new critical path. The specific implementation is as follows: First, a dynamic resource dependency network is constructed. The nodes of the dynamic resource dependency network consist of all the pending tasks that have been affected by the cascading effects, with each pending task being a task node in the network. The edges of the dynamic resource dependency network represent the interdependencies between task nodes caused by the cascading effects. These interdependencies are determined and established according to three specific rules. The first rule is based on dependencies on shared critical conflicting resources. If two different pending tasks are both identified as depending on the same critical conflicting resource in step S3, then a connection edge is established between these two task nodes, indicating that there is a competitive dependency between them on the same scarce resource. The second rule is based on the dependency relationship of overlapping planned task time window offsets. For each task to be executed, the adjusted new planned task time window can be determined based on its planned task time window offset calculated in step S3. If the adjusted new planned task time windows of any two tasks to be executed overlap in time, a connection edge is established between the two task nodes, indicating that they have a conflict or correlation in time scheduling. The criterion for overlap is that the intersection of the two time windows on the time axis is not empty, that is, the start time of one window is earlier than the end time of the other window, and its end time is later than the start time of the other window. The third rule is based on dependencies with conflicting additional energy replenishment needs. If two different tasks have additional energy replenishment needs calculated in step S3 that both point to the same alternative energy supply facility, and the planned time windows of these two tasks are expected to arrive at the alternative facility simultaneously or at approximately the same time after adjustment, then their additional energy replenishment needs are considered to conflict, and a connection edge is established between the two task nodes. The criterion for determining simultaneous or approximately the same time is that the time difference between the expected arrival times of the two tasks at the alternative facility is less than a preset time conflict threshold. The time conflict threshold is obtained based on historical data analysis or operational experience. For example, by analyzing historical data on vehicle queuing at charging stations, the upper limit of the queuing time that vehicles can generally accept is statistically determined, and the statistical average or median of this upper limit is used as the basis for the time conflict threshold. The specific value is, for example, 15 minutes. By traversing all task node pairs through the above rules, a network containing all nodes and dependent edges is established, which is a dynamic resource dependency network. Its topology changes according to the specific content of the chain reaction.

[0031] Secondly, the betweenness centrality of each task node in the dynamic resource dependency network before and after the change in inter-task dependencies caused by cascading effects is calculated, and a sequence of betweenness centrality change values ​​is obtained. This calculation process is divided into two contrasting states: the original network state before the cascading effects and the new network state after the cascading effects. For the original network state before the cascading effects, its network topology is constructed based on the normal task planning relationships before the statistically significant mutation was detected in step S2. At this time, there are usually only a few dependency edges between task nodes due to the preset execution order or geographical proximity. For the new network state after the cascading effects, its network topology is the dynamic resource dependency network just constructed in the previous step. For each network state, the shortest path between each pair of task nodes in the network under that state is calculated. The shortest path is the path that connects two task nodes with the fewest number of edges. Dependency edges are assigned the same weight value by default when calculating the shortest path, that is, the weight of each edge is 1. The shortest path is the path with the fewest number of edges. The calculation of the shortest path is implemented using classic graph theory algorithms, such as the breadth-first search algorithm. For each task node in the network, count the number of shortest paths among all node pairs that pass through the current task node in that network state. This number is the shortest path count for that node. Simultaneously, count the total number of shortest paths between all task node pairs in that network state, excluding paths with the same origin and destination node. The betweenness centrality of a task node in a given network state is defined as the number of shortest paths passing through that node divided by the total number of shortest paths between all task node pairs in that state; the quotient is a value between 0 and 1. Calculate the betweenness centrality of each task node in both the state before and after a cascading effect occurs. Subtract the betweenness centrality value of the node before the cascading effect from the betweenness centrality value in the state after the cascading effect occurs to obtain the change in betweenness centrality value for that node. Repeat the above calculation at fixed time or event intervals, such as after each new cascading effect is detected, to form a sequence of betweenness centrality change values ​​arranged in chronological order for each task node.

[0032] Next, the Mann-Kendall trend test algorithm was used to analyze the betweenness centrality change value sequence of each task node, identifying consecutive task nodes with a statistically significant upward trend in betweenness centrality. The Mann-Kendall trend test is a non-parametric statistical test method used to determine whether a sequence of data exhibits a monotonically increasing or decreasing trend. For the betweenness centrality change value sequence of each task node, all data points in the sequence are arranged in chronological order. The Mann-Kendall statistic for this sequence is calculated by comparing each data point in the sequence with all subsequent data points, recording the number of pairs where the subsequent value is greater than the previous value, and then recording the number of pairs where the subsequent value is less than the previous value. The total number of the former minus the total number of the latter yields an initial statistic. Based on the length of the sequence, this initial statistic is standardized to obtain a standard normal distribution statistic. A significance level threshold is preset. Setting this threshold is a standard step in statistical hypothesis testing, used to control the probability of falsely rejecting the null hypothesis. A common value is, for example, 0.05. By comparing the p-value corresponding to the calculated standard normal distribution statistic with the preset significance level threshold, it is determined whether the sequence exhibits a statistically significant trend. If the calculated p-value is less than the preset significance level threshold and the standardized statistic is positive, then the sequence of betweenness centrality changes of the task node is considered to have a statistically significant upward trend. Continuous task nodes refer to a group of task nodes that are directly connected in the topology by dependency edges and whose betweenness centrality change sequences are all identified as exhibiting a statistically significant upward trend. The identification method is to traverse all connecting edges in the dynamic resource dependency network. If two task nodes connected by an edge both satisfy the above trend test conditions, then these two nodes are considered continuous nodes. By connecting edges that meet the conditions, the longest chain or several chains composed of such nodes can be found in the network.

[0033] Finally, the identified consecutive task nodes are connected sequentially according to their corresponding planned operation time windows to form a new critical path. For each identified task chain consisting of consecutive task nodes, the new planned operation time window for each task node in the chain, after adjustment in step S3, is obtained, especially the start time of the window. The nodes in the task chain are sorted according to the time order from earliest to latest starting time. The task nodes are then connected sequentially from the first to the last according to this time order, and the resulting path is the new critical path. This new critical path represents a bottleneck sequence consisting of a series of tasks that are tightly coupled in terms of resource dependence and time scheduling and whose network importance increases synchronously under the cascading effect. Its stability directly affects the execution efficiency of the entire fleet's subsequent scheduling plan.

[0034] S5. Based on the new critical path, by weighing the combined impact of different scheduling adjustment strategies on the operational and energy efficiency of the new critical path, evaluate the long-term comprehensive effectiveness of each scheduling adjustment strategy. Specifically, the implementation is as follows: First, for the new critical path, multiple scheduling adjustment strategies are generated. These strategies include at least three basic types: adjusting task execution order, adjusting critical conflict resource allocation, and inserting preset charging periods for tasks. The specific method for generating a task execution order adjustment strategy is to list all possible permutations and combinations of the multiple tasks to be executed in the new critical path, without violating the inherent logical dependencies between tasks. Each different task execution sequence constitutes a task execution order adjustment strategy. The specific method for generating a critical conflict resource allocation adjustment strategy is to list all alternative energy supply facilities for each task in the new critical path that depends on critical conflict resources. Alternative energy supply facilities refer to energy supply facilities, besides the originally planned critical conflict resources, that are within their preset third correlation distance threshold and whose real-time available output power can meet the task's requirements. Allocating different combinations of alternative facilities to the task constitutes different critical conflict resource allocation adjustment strategies. The specific method for generating a strategy to insert a preset charging period is as follows: Between any two consecutive task nodes in the new critical path, or at the beginning and end of the path, a charging period of preset duration is planned and inserted. The preset duration is a time threshold parameter, which is obtained based on the typical charging speed of the sanitation vehicle's power battery and operational experience. For example, by statistically analyzing historical charging records, an average charging time sufficient to replenish the typical amount of electricity required for the vehicle to perform critical path tasks is obtained. This average charging time is used as the basis for determining the preset duration, for example, 30 minutes. A specific energy supply facility is then assigned to this charging period. Different combinations of insertion locations, charging durations, and charging facilities constitute different strategies for inserting preset charging periods. Each specific scheduling adjustment strategy can be composed of one or more of the above basic adjustment operations.

[0035] The third association distance threshold is a pre-set distance value used to determine whether a task to be performed or a sanitation vehicle has a plannable spatial association with a certain energy supply facility. The method for obtaining and setting this third association distance threshold is as follows: its setting is mainly based on the typical driving range of sanitation vehicles after a single charge, the spatial distribution characteristics of the work area, and operational experience. First, by analyzing the fleet's historical operational data, the average driving distance of vehicles before and after completing typical tasks to charging facilities is statistically analyzed, and its distribution pattern is calculated; for example, the 85th percentile driving distance is used as a reference benchmark. Second, combined with the current layout density of all energy supply facilities in the network, it is ensured that the set threshold ensures that the vast majority of task locations or vehicle positions can be associated with at least one alternative facility within its range. For example, based on the above analysis, the third association distance threshold can be set to 3000 meters. This means that when generating and adjusting the allocation strategy for critical conflict resources, only other energy supply facilities located within 3000 meters of the original critical conflict resource facility will be listed as alternative facilities. The preset third correlation distance threshold range refers to a circular geographical area centered on a specific facility location and with this third correlation distance threshold as its radius. Therefore, when determining whether an energy supply facility is a potential candidate, the Euclidean distance between the facility's coordinates and the task location coordinates or the vehicle's real-time location coordinates is calculated. This distance is then compared to the third correlation distance threshold. If the distance is less than or equal to the third correlation distance threshold, the facility is considered to be within the third correlation distance threshold range. The specific value of the third correlation distance threshold can be configured differently based on the facility layout in different urban areas.

[0036] Secondly, based on real-time vehicle status data, real-time task demand data, and the additional energy replenishment requirements and planned operation time window offsets in the chain reaction, the cumulative operation time and cumulative energy consumption required to complete all tasks in the new critical path after executing each scheduling adjustment strategy are simulated. The simulation execution requires setting uniform initial conditions and environmental parameters. The initial conditions for the simulation are the real-time vehicle status data at the start of the scheduling adjustment, including the real-time location information, real-time driving speed, and real-time power battery state of charge data of the relevant sanitation vehicles. The simulation execution process is based on the logical progression of discrete event simulation. For each scheduling adjustment strategy to be evaluated, the complete process of sanitation vehicles performing tasks is simulated in a time-step manner according to the new task sequence, new resource allocation scheme, and inserted charging period arrangement specified by the strategy. In each time step, the energy consumed and operation time spent on the task are calculated based on the estimated workload data in the real-time task demand data and the vehicle's unit operation energy consumption efficiency parameter. Simultaneously, the travel distance is calculated based on the changes in the vehicle's real-time location information and the target location, and combined with the vehicle's unit distance travel energy consumption parameter, the travel energy consumption and travel time are calculated. When the simulation reaches the preset charging period, the increased energy replenishment and charging time during the charging process are simulated based on the real-time available output power data of the charging facility and the charging characteristic curve of the vehicle battery. The offset of the planned operation time window in the cascading effect participates in the simulation as a time boundary constraint for the allowed execution of tasks, ensuring that the task execution time in the simulation falls within the adjusted planned operation time window. The entire simulation process continues until the last task in the new critical path is completed according to the strategy definition. After the simulation, the total time spent on all task executions under this strategy is summarized to obtain the cumulative operation time, which is in hours; the total electrical energy consumed by all task executions, vehicle movement, and charging processes under this strategy is summarized to obtain the cumulative energy consumption, which is in kilowatt-hours.

[0037] Finally, for each scheduling adjustment strategy, based on the simulated cumulative operation time and cumulative energy consumption, combined with preset operation time weighting coefficients and energy consumption weighting coefficients, a strategy evaluation value representing long-term comprehensive effectiveness is calculated. The specific methods for presetting the operation time weighting coefficient and energy consumption weighting coefficient are as follows: These two coefficients are used to quantify the relative importance of operation efficiency targets and energy efficiency targets in the comprehensive evaluation. Their values ​​are set based on a process of historical data analysis and operational strategy trade-offs. Specifically, the acquisition and setting methods include collecting and analyzing the actual operation time data and actual energy consumption data of all tasks performed by the fleet within a complete past operating cycle, such as the past month; calculating the additional costs caused by operation time delays and the costs caused by energy consumption respectively; analyzing the proportion of each in the total additional costs; and normalizing these two cost proportions as the basis for setting the operation time weighting coefficient and energy consumption weighting coefficient. The weighting coefficients are the direct basis for determining the operational time delay cost and the energy consumption cost cost. For example, if analysis shows that the operational time delay cost accounts for 60% and the energy consumption cost accounts for 40%, then the operational time weighting coefficient can be set to 0.6 and the energy consumption weighting coefficient to 0.4. Alternatively, the operational management personnel can directly decide on the weighting based on the clear management objectives of the current stage. For example, during the morning peak hours, ensuring the speed of road cleaning is the priority, so operational time is given a higher weight; during the afternoon off-peak hours, reducing operating costs is the priority, so energy consumption is given a higher weight. Regardless of the method used, it must be ensured that both the operational time weighting coefficient and the energy consumption weighting coefficient are greater than 0, and their sum equals 1. After obtaining the weighting coefficients, the specific steps for calculating the strategy evaluation value are as follows. First, the cumulative operation time data of all strategies obtained through simulation are standardized to eliminate the influence of dimensions. Standardization methods include identifying the minimum and maximum cumulative operation time among all strategies. For a given strategy, the minimum cumulative operation time is subtracted from the cumulative operation time, and then divided by the difference between the maximum and minimum cumulative operation time, resulting in a standardized value between 0 and 1. The cumulative energy consumption data is processed using the same standardization method. Then, for each scheduling adjustment strategy, its standardized cumulative operation time value is multiplied by a preset operation time weighting coefficient to obtain a weighted operation time score; its standardized cumulative energy consumption value is multiplied by a preset energy consumption weighting coefficient to obtain a weighted energy consumption score. Finally, the weighted operation time score and weighted energy consumption score of the strategy are added together, and the sum is the strategy evaluation value. The strategy evaluation value is a dimensionless value; the smaller the value, the better the long-term comprehensive performance of the strategy in balancing operation efficiency and energy efficiency. Through the above calculations, a comparable policy evaluation value is calculated for each generated scheduling adjustment policy.

[0038] S6. Generate and issue target collaborative scheduling instructions based on long-term comprehensive efficiency to optimize the overall long-term comprehensive efficiency of the fleet. The specific implementation is as follows: First, compare the strategy evaluation values ​​corresponding to all scheduling adjustment strategies, and determine the scheduling adjustment strategy with the optimal strategy evaluation value as the target scheduling adjustment strategy. The strategy evaluation value is a dimensionless numerical value characterizing long-term overall effectiveness; the smaller the value, the better the long-term overall effectiveness of the scheduling adjustment strategy in balancing operational efficiency and energy efficiency. Therefore, the specific method for determining the optimal strategy evaluation value is as follows: traverse all scheduling adjustment strategies generated and evaluated for the new critical path in step S5, obtain the strategy evaluation value corresponding to each scheduling adjustment strategy, and find the one with the smallest value among all strategy evaluation values. The scheduling adjustment strategy corresponding to the strategy with the smallest value is determined as the target scheduling adjustment strategy. During the comparison process, if two or more scheduling adjustment strategies have the same strategy evaluation value and are all minimum values, additional decision rules can be used for final determination. For example, prioritize the strategy with shorter cumulative operation time or prioritize the strategy with lower cumulative energy consumption. The activation and specific selection logic of this additional decision rule can be preset according to actual operational needs.

[0039] Secondly, based on the operations included in the target scheduling adjustment strategy, such as adjusting the task execution order, adjusting the allocation of critical conflict resources, or inserting preset charging periods, a target collaborative scheduling instruction containing specific execution times, target resources, and action parameters is generated. The process of generating instructions is a step of transforming the abstract scheme of the target scheduling adjustment strategy into concrete, executable commands. First, the specific adjustment operation types and contents included in the target scheduling adjustment strategy are analyzed. If the strategy includes operations to adjust the task execution order, a clear task sequence instruction needs to be generated. Specifically, according to the new task sequence defined by the strategy, an execution sequence number is assigned to each task to be executed in the sequence, and the estimated start and end times of each task based on the new sequence are calculated. The estimated start time is calculated based on the estimated end time of the preceding task, vehicle travel time, and the adjusted planned operation time window of the task itself; this time is the specific execution time point or time window that needs to be included in the instruction. If the strategy includes operations to adjust the allocation of critical conflict resources, a resource reallocation instruction needs to be generated. Specifically, for tasks requiring changes to energy supply facilities as specified in the strategy, the identifier of the new target energy supply facility, such as the charging pile number, is clearly defined. The optimal route and estimated arrival time for the vehicle to reach this facility are calculated. This estimated arrival time, along with the facility identifier, constitutes the key parameters of the resource allocation instruction. If the strategy includes inserting a preset charging period, a charging task instruction must be generated. Specifically, the start time, duration, and designated charging facility identifier of the inserted charging period are specified. The charging start time is determined based on the connection time between preceding and following tasks and the vehicle's battery status, while the duration uses a preset charging duration threshold from the strategy. Finally, all the sub-instructions generated for different operations are integrated and encoded chronologically to form a complete, structured target collaborative scheduling instruction. This instruction can be a data message containing a timestamp, vehicle identifier, task identifier, action type code, and parameter list.

[0040] The specific method for obtaining and setting the charging time threshold is as follows: Its setting is mainly based on the charging characteristics of the power battery configured in the sanitation vehicle, the typical output power of commonly used charging facilities, and operational efficiency considerations. First, based on the model and technical parameters of the vehicle's power battery, the nominal charging time required to charge from a typical low-charge state to a safe charge state sufficient for subsequent tasks is obtained. This nominal charging time can be calculated using charging curves and capacity data provided by the battery manufacturer. Second, considering the rated output power range of commonly used charging facilities in the fleet, such as DC fast charging piles, and taking into account uncertainties such as power fluctuations and queuing during actual operation, a buffer time margin is added to the nominal charging time to form a more robust charging time estimate. For example, based on the above analysis, if a certain model of sanitation vehicle requires 45 minutes to charge from 30% to 80% at rated power, the charging time threshold can be set to 60 minutes after considering the buffer. The charging time threshold will serve as the specific numerical basis for determining the preset duration or continuous duration when generating strategies for inserting preset charging periods and subsequent charging task instructions. The charging time threshold can be configured differently depending on the vehicle model or the type of charging facility.

[0041] Finally, the target coordinated dispatch instruction is issued to the corresponding sanitation vehicles and energy supply facilities. This issuance process requires a reliable communication link and a clearly defined recipient address. For sanitation vehicles, instruction issuance is based on vehicle-to-everything (V2X) communication technology. Specifically, based on the sanitation vehicle identifier in the instruction, the data packet of the target coordinated dispatch instruction is sent to the corresponding vehicle's onboard IoT terminal via a wireless communication network, such as a 4G or 5G network. Upon receiving the instruction, the onboard IoT terminal parses and verifies it, converting it into display information and control signals that can be recognized by the vehicle's infotainment system or control unit, guiding the driver or allowing the vehicle with some autonomous driving capabilities to execute it autonomously. For energy supply facilities, instruction issuance typically aims to notify the facility management system to pre-allocate or reserve status for upcoming vehicle service requests. Specifically, based on the energy supply facility identifier in the instruction, the facility-related portion of the instruction, such as the vehicle's estimated arrival time and charging power requirement, is issued to the facility's controller or local management system via the IoT platform. The facility's system updates its resource reservation status according to the instruction and prioritizes service for the vehicle upon arrival. To ensure the reliability and timeliness of command issuance, a communication mechanism with acknowledgment and retransmission can be adopted. This means that after sending a command, the system waits for an acknowledgment from the receiver. If no acknowledgment is received within a preset timeout threshold, the command is automatically retransmitted. The timeout threshold is set based on the average round-trip time and reliability requirements of the communication network. It is determined by obtaining average latency data through multiple network communication tests and adding a safety margin. For example, if the average latency is 1 second and the safety margin is set to 2 seconds, then the timeout threshold is set to 3 seconds. This process completes a closed loop from policy decision-making to final command generation and issuance, enabling the optimal scheduling adjustment scheme derived from long-term comprehensive performance evaluation to be implemented in actual operation.

[0042] Example 2: Figure 2 A schematic diagram of the IoT-based intelligent sanitation vehicle collaborative dispatching system of the present invention is provided. The IoT-based intelligent sanitation vehicle collaborative dispatching system includes: The data acquisition module is used to acquire real-time vehicle status data of sanitation vehicles, real-time task requirement data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. The mutation identification module is used to identify resource competition networks and assess whether statistically significant mutations have occurred in the rate of change of their topological vulnerability indicators based on real-time vehicle status data, real-time task requirement data, and real-time energy status data. The cascading analysis module is used to analyze the cascading effects on subsequent tasks of other vehicles in the fleet when a statistically significant mutation occurs. The path determination module is used to construct a dynamic resource dependency network based on cascading effects, identify consecutive task nodes that have accumulated mutations in betweenness centrality due to cascading effects, and determine the task chain formed by them as a new critical path. The performance evaluation module is used to evaluate the long-term overall performance of each scheduling adjustment strategy based on the new critical path by weighing the combined impact of different scheduling adjustment strategies on the operational efficiency and energy efficiency of the new critical path. The instruction issuance module is used to generate and issue target collaborative scheduling instructions that optimize the overall long-term comprehensive efficiency of the fleet based on long-term comprehensive performance.

[0043] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0044] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0045] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. Computer-readable storage media can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0046] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0047] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0048] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0049] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0050] If a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0051] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0052] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart sanitation vehicle collaborative scheduling method based on the Internet of Things, characterized in that, include: S1. Obtain real-time vehicle status data of sanitation vehicles, real-time task demand data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. S2. Based on real-time vehicle status data, real-time task requirement data, and real-time energy status data, identify resource competition networks and assess whether their topological vulnerability index change rate has undergone statistically significant abrupt changes. S3. When a statistically significant abrupt change occurs, analyze the resulting cascading effects on the subsequent tasks of other vehicles in the convoy; S4. Construct a dynamic resource dependency network based on cascading effects, and identify consecutive task nodes that have undergone betweenness centrality cumulative mutations due to cascading effects, and determine the task chain formed by them as the new critical path. S5. Based on the new critical path, evaluate the long-term comprehensive effectiveness of each scheduling adjustment strategy by weighing the combined impact of different scheduling adjustment strategies on the operational efficiency and energy efficiency of the new critical path. S6. Generate and issue target collaborative scheduling instructions based on long-term comprehensive performance to optimize the overall long-term comprehensive performance of the fleet.

2. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 1, characterized in that, S1 includes: The real-time location information, real-time driving speed and real-time power battery state of charge data of sanitation vehicles are obtained as real-time vehicle status data. Obtain the task location coordinates, planned task time window, and estimated task load data of the tasks to be executed as real-time task requirement data; The location coordinates, real-time occupancy status, and real-time available output power data of energy supply facilities are obtained as real-time energy status data.

3. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 1, characterized in that, S2 include: Taking sanitation vehicles and tasks to be performed as the first type of nodes, and energy supply facilities and high-conflict road sections as the second type of nodes, a weighted connection edge reflecting the demand of vehicles and tasks for the second type of nodes is established based on the task location coordinates and planned operation time window in the real-time task demand data, and the real-time location information in the real-time vehicle status data, thereby constructing a resource competition network. The spectral radius of the resource contention network over a continuous time window is calculated as a topological vulnerability index, and the rate of change of the topological vulnerability index over time is obtained. The cumulative sum control chart algorithm is used to monitor the rate of change of topological vulnerability indicators over time to determine whether there are statistically significant abrupt changes in the rate of change of topological vulnerability indicators.

4. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 3, characterized in that, The spectral radius of a resource-competing network over a continuous time window is used as a topological vulnerability indicator. This is achieved by representing the constructed resource-competing network as a weighted adjacency matrix, where the element values ​​are determined by the weights of the corresponding connecting edges. For each continuous time window, the weighted adjacency matrix is ​​used to calculate all its eigenvalues ​​and take their absolute values. The eigenvalue with the largest absolute value is the spectral radius of the resource-competing network for that time window, which is then used as the topological vulnerability indicator.

5. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 1, characterized in that, S3 include: When a statistically significant mutation occurs, based on the resource competition network that has been identified as having a statistically significant mutation, energy supply facilities whose demand exceeds the real-time available output power due to changes in the weight of the connecting edges in the resource competition network are identified as key conflict resources. Determining all planned job time windows depends on pending job tasks that use critical conflicting resources; For each pending task that depends on critical conflict resources, the additional energy replenishment requirement and the offset of the planned operation time window are calculated based on the real-time power battery state of charge data in the real-time vehicle status data and the estimated operation load data in the real-time task demand data. The additional energy replenishment requirements and planned operation time window offsets calculated for all critical conflict resources are defined as cascading effects.

6. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 1, characterized in that, S4 include: A dynamic resource dependency network is constructed using the tasks to be executed as nodes, as well as the dependencies between tasks caused by sharing key conflicting resources, overlapping planned operation time window offsets, or conflicting additional energy replenishment needs. Calculate the betweenness centrality of each task node in the dynamic resource dependency network before and after the change in inter-task dependency caused by cascading effects, and obtain its sequence of betweenness centrality change values. The Mann-Kendall trend test algorithm was used to analyze the sequence of betweenness centrality changes for each task node, and consecutive task nodes with statistically significant upward trends in betweenness centrality were identified. The identified consecutive task nodes are connected sequentially according to their corresponding planned operation time windows to form a new critical path.

7. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 6, characterized in that, The betweenness centrality of each task node in a dynamic resource dependency network before and after the change in inter-task dependencies caused by cascading effects is calculated as follows: Based on the topology of the constructed dynamic resource dependency network, the shortest path between each pair of task nodes in the network is calculated in both the network states before and after the cascading effects occur; for each task node, the number of shortest paths passing through the corresponding task node in each state is counted, and the ratio of the number of shortest paths to the total number of shortest paths for all pairs of task nodes in the network in the corresponding state is defined as the betweenness centrality of the corresponding task node in this state.

8. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 1, characterized in that, S5 include: For the new critical path, generate at least a variety of scheduling adjustment strategies, including adjusting the task execution order, adjusting the allocation of critical conflict resources, and inserting preset charging periods for tasks; Based on real-time vehicle status data, real-time task demand data, and additional energy replenishment demand and planned operation time window offset in the chain reaction, the cumulative operation time and cumulative energy consumption required to complete all tasks in the new critical path after each scheduling adjustment strategy is simulated. For each scheduling adjustment strategy, based on the simulated cumulative operation time and cumulative energy consumption, combined with the preset operation time weight coefficient and energy consumption weight coefficient, a strategy evaluation value representing the long-term comprehensive effectiveness is calculated.

9. The IoT-based intelligent sanitation vehicle collaborative scheduling method according to claim 1, characterized in that, S6 include: Compare the policy evaluation values ​​corresponding to all scheduling adjustment policies, and determine the scheduling adjustment policy with the best policy evaluation value as the target scheduling adjustment policy; Based on the target scheduling adjustment strategy, which includes operations such as adjusting the task execution order, adjusting the allocation of key conflicting resources, or inserting preset charging periods, a target collaborative scheduling instruction containing specific execution time, target resources, and action parameters is generated. The target coordinated dispatch instructions are issued to the corresponding sanitation vehicles and energy supply facilities.

10. An IoT-based intelligent sanitation vehicle collaborative scheduling system, used to implement the IoT-based intelligent sanitation vehicle collaborative scheduling method according to any one of claims 1-9, characterized in that, include: The data acquisition module is used to acquire real-time vehicle status data of sanitation vehicles, real-time task requirement data of tasks to be performed, and real-time energy status data of energy supply facilities through the Internet of Things. The mutation identification module is used to identify resource competition networks and assess whether statistically significant mutations have occurred in the rate of change of their topological vulnerability indicators based on real-time vehicle status data, real-time task requirement data, and real-time energy status data. The cascading analysis module is used to analyze the cascading effects on subsequent tasks of other vehicles in the fleet when a statistically significant mutation occurs. The path determination module is used to construct a dynamic resource dependency network based on cascading effects, identify consecutive task nodes that have accumulated mutations in betweenness centrality due to cascading effects, and determine the task chain formed by them as a new critical path. The performance evaluation module is used to evaluate the long-term overall performance of each scheduling adjustment strategy based on the new critical path by weighing the combined impact of different scheduling adjustment strategies on the operational efficiency and energy efficiency of the new critical path. The instruction issuance module is used to generate and issue target collaborative scheduling instructions that optimize the overall long-term comprehensive efficiency of the fleet based on long-term comprehensive performance.

Images