Cloud-edge collaborative load scheduling method, apparatus and electronic equipment

Abstract

This invention discloses a method, apparatus, and electronic device for load equipment scheduling based on cloud-edge collaboration. The method includes: determining the equipment data of the load equipment corresponding to each edge device; clustering the equipment data of each load device to obtain multiple data sets; determining the target scheduling tasks corresponding to each of the multiple data sets; uploading the multiple target scheduling tasks to the cloud for processing to obtain cloud scheduling parameters corresponding to each target scheduling task; determining the matching index between each edge device and the multiple target scheduling tasks; and determining the target scheduling parameters for the load equipment corresponding to each edge device based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks. This invention solves the technical problem of high scheduling costs in related technologies when scheduling load equipment based on edge devices and cloud collaboration.

Description

Technical Field

[0001] This invention relates to the field of power systems, and more specifically, to a load equipment scheduling method, apparatus, and electronic equipment based on cloud-edge collaboration. Background Technology

[0002] In related technologies, scheduling load devices through cloud collaboration can achieve global unified control of a large number of load devices. However, in related technologies, when scheduling load devices based on edge devices and cloud collaboration, there is a technical problem of high scheduling costs.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a method, apparatus, and electronic device for scheduling load devices based on cloud-edge collaboration, in order to at least solve the technical problem of high scheduling costs in related technologies when scheduling load devices based on collaborative scheduling between edge devices and the cloud.

[0005] According to one aspect of the present invention, a load device scheduling method based on cloud-edge collaboration is provided, comprising: determining device data of load devices corresponding to each edge device; clustering the device data of the load devices corresponding to each edge device to obtain multiple data sets; determining target scheduling tasks corresponding to the multiple data sets; uploading the multiple target scheduling tasks to the cloud for processing to obtain cloud scheduling parameters corresponding to the multiple target scheduling tasks; determining a matching index between each edge device and the multiple target scheduling tasks, wherein the matching index is used to characterize the degree of matching between the device scheduling task corresponding to the corresponding edge device and the corresponding target scheduling task; and determining target scheduling parameters of the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks, for scheduling the corresponding load devices.

[0006] Optionally, the step of uploading multiple target scheduling tasks to the cloud for processing to obtain cloud scheduling parameters corresponding to the multiple target scheduling tasks includes: determining the uplink and downlink transmission speeds of the edge devices corresponding to the multiple target scheduling tasks, wherein the uplink transmission speed is the data transmission speed when the corresponding edge device uploads data to the cloud, and the downlink transmission speed is the data transmission speed when the cloud sends data to the corresponding edge device; determining the cloud computing power configuration corresponding to the multiple target scheduling tasks based on the task size and predicted size corresponding to the multiple target scheduling tasks, and the uplink and downlink transmission speeds of the edge devices corresponding to the multiple target scheduling tasks, wherein the predicted size is the predicted data size of the task processing result after the corresponding target scheduling task is processed; and processing the target scheduling tasks corresponding to the multiple target scheduling tasks based on the cloud computing power configuration corresponding to the multiple target scheduling tasks to obtain the cloud scheduling parameters corresponding to the multiple target scheduling tasks.

[0007] Optionally, determining the cloud computing power configuration corresponding to each of the multiple target scheduling tasks based on the task size and predicted size corresponding to each of the multiple target scheduling tasks, and the uplink and downlink transmission speeds of the edge devices corresponding to each of the multiple target scheduling tasks, includes: calling a target function, wherein the target function is a function aimed at minimizing task processing time, wherein the target function includes an uplink transmission time item, a downlink transmission time item, and a data processing time item, wherein the uplink transmission time item is the item corresponding to the data transmission time when the corresponding edge device uploads data to the cloud, and the downlink transmission time item is the item corresponding to the data transmission time when the cloud sends data to the corresponding edge device; and determining the cloud computing power configuration corresponding to each of the multiple target scheduling tasks based on the target function, the task size and predicted size corresponding to each of the multiple target scheduling tasks, and the uplink and downlink transmission speeds of the edge devices corresponding to each of the multiple target scheduling tasks.

[0008] Optionally, when the plurality of target scheduling tasks each include a plurality of subtasks, the objective function is expressed as:

[0009]

[0010] in, The task processing time corresponding to the target scheduling task; This represents the total number of edge devices corresponding to the target scheduling task. The total number of subtasks corresponding to the target scheduling task; For the first The first target scheduling task corresponding to the edge device The size of each subtask; For the first The first target scheduling task corresponding to the edge device Sub-computing power configuration for each sub-task; For the first Uplink transmission speed of each edge device; For the first Downlink transmission speed of individual edge devices; For the first The first target scheduling task corresponding to the edge device Sub-prediction size of each sub-task.

[0011] Optionally, determining the target scheduling parameters of the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the plurality of target scheduling tasks and the matching index between each edge device and the plurality of target scheduling tasks includes: determining the reference scheduling range of the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the plurality of target scheduling tasks and the matching index between each edge device and the plurality of target scheduling tasks; determining the operation and maintenance cost and power interaction cost of the load devices corresponding to each edge device, wherein the power interaction cost is the cost incurred when the load devices corresponding to the edge devices interact with the power grid; and determining the target scheduling parameters of the load devices corresponding to each edge device based on the reference scheduling range, operation and maintenance cost, and power interaction cost of the load devices corresponding to each edge device.

[0012] Optionally, determining the target scheduling parameters for the load devices corresponding to each edge device based on the reference scheduling range, operation and maintenance cost, and power interaction cost of the load devices corresponding to each edge device includes: determining the state change constraints of the load devices corresponding to each edge device; and determining the target scheduling parameters for the load devices corresponding to each edge device based on the state change constraints, reference scheduling range, operation and maintenance cost, and power interaction cost of the load devices corresponding to each edge device.

[0013] Optionally, the step of clustering the device data of the load devices corresponding to each edge device to obtain multiple data sets includes: determining the device type of the load devices corresponding to each edge device, as well as the data type and data size of the device data of the load devices corresponding to each edge device; and clustering the device data of the load devices corresponding to each edge device based on the device type of the load devices corresponding to each edge device, as well as the data type and data size of the device data of the load devices corresponding to each edge device, to obtain multiple data sets.

[0014] According to one aspect of the present invention, a load device scheduling device based on cloud-edge collaboration is provided, comprising: a first determining module, configured to determine device data of load devices corresponding to each edge device; a second determining module, configured to cluster the device data of the load devices corresponding to each edge device to obtain multiple data sets; a third determining module, configured to determine target scheduling tasks corresponding to the multiple data sets; a fourth determining module, configured to upload the multiple target scheduling tasks to the cloud for processing to obtain cloud scheduling parameters corresponding to the multiple target scheduling tasks; a fifth determining module, configured to determine a matching index between each edge device and the multiple target scheduling tasks, wherein the matching index is used to characterize the degree of matching between the device scheduling task corresponding to the corresponding edge device and the corresponding target scheduling task; and a sixth determining module, configured to determine target scheduling parameters of the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks, for scheduling the corresponding load devices.

[0015] According to one aspect of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the cloud-edge collaborative load scheduling method described in any of the preceding claims.

[0016] According to one aspect of the present invention, a computer-readable storage medium is provided, wherein when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the cloud-edge collaborative load scheduling method described above.

[0017] In this embodiment of the invention, by clustering the load data of each edge device into multiple data sets and matching each set with a corresponding target scheduling task, dispersed heterogeneous data can be integrated into representative task units and uploaded to the cloud, thereby reducing data redundancy and computational load in cloud processing. Then, based on the scheduling parameters returned by the cloud to each task unit, and combined with the matching index between the edge device and the task, the target scheduling parameters of each edge device are allocated. Since the matching index accurately reflects the degree of fit between the device scheduling task and the target scheduling task, the global optimization parameters generated by the cloud can be adapted to each edge device in a differentiated manner, avoiding the duplicate calculations and invalid communication caused by all edge devices directly using the same cloud parameters. This solves the technical problem of high scheduling cost in related technologies when coordinating load devices with the edge device and the cloud. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0019] Figure 1 This is a flowchart of a cloud-edge collaborative load scheduling method according to an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of a load device scheduling framework based on cloud-edge collaboration in an optional embodiment of the present invention;

[0021] Figure 3 This is a structural block diagram of a cloud-edge collaborative load scheduling device according to an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] Example 1

[0025] According to an embodiment of the present invention, an embodiment of a load device scheduling method based on cloud-edge collaboration is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0026] Figure 1 This is a flowchart of a cloud-edge collaborative load scheduling method according to an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes the following steps:

[0027] S102, determine the equipment data of the load equipment corresponding to each edge device;

[0028] This involves edge devices, which are localized computing and processing units deployed on the power user side or near the load end. For example, the edge devices include edge computing gateways, intelligent data concentrators, and area controllers, which are used to collect, preprocess, perform preliminary analysis, and local control of data from load devices within the control range / corresponding to the local data, so as to reduce the computing and communication pressure on the cloud data center.

[0029] This involves load equipment, which is an adjustable power load device on the end-user side. For example, the load equipment includes various controllable loads such as electric vehicles, air conditioning loads, distributed electric heating equipment, energy storage equipment, and heat pump loads. It is used to respond to dispatch instructions and participate in the supply and demand balance and coordinated interaction of the power grid by adjusting its own operating status.

[0030] This involves equipment data, which is data that characterizes the operating status, power consumption characteristics, and environmental parameters of the load equipment. For example, the equipment data includes real-time power information, voltage and current data, equipment operating time, temperature data, charging and discharging power, and equipment type identification of the load equipment.

[0031] By determining the device data of the load devices corresponding to each edge device, it is possible to collect multi-source heterogeneous information scattered on the user side in a structured manner locally, providing a data foundation for subsequent processing.

[0032] S104, cluster the equipment data of the load devices corresponding to each edge device to obtain multiple datasets;

[0033] Clustering the load device data corresponding to each edge device and forming multiple data sets can organize scattered and heterogeneous data into unified units according to features. Since clustering can eliminate data redundancy and disorder, it can reduce the computing and transmission burden of cloud processing, solve the problem of low efficiency caused by complex data in cloud-edge collaborative scheduling, and provide a standardized data foundation for subsequent scheduling task matching and parameter distribution.

[0034] Optionally, the equipment data of the load devices corresponding to each edge device are clustered to obtain multiple datasets, including: determining the equipment type of the load devices corresponding to each edge device, as well as the data type and data scale of the equipment data of the load devices corresponding to each edge device; and clustering the equipment data of the load devices corresponding to each edge device based on the equipment type of the load devices corresponding to each edge device, as well as the data type and data scale of the equipment data of the load devices corresponding to each edge device, to obtain multiple datasets.

[0035] This involves data clustering, which is the process of grouping and classifying device data according to preset similarity rules. For example, data clustering can be performed based on device type, data type, and data size so that data with the same or similar attributes can be integrated into the same set.

[0036] This involves a dataset, which consists of clustered data groups of similar types. Each group contains load equipment data with similar characteristics and is used as a processing unit for unified scheduling tasks.

[0037] This involves equipment type, which is the category of load equipment. For example, equipment types include electric vehicles, air conditioners, energy storage, etc., which are used to distinguish the operating characteristics and control requirements of different loads.

[0038] This involves data types, which are the attribute categories of the data collected by the load equipment, such as power data, voltage data, temperature data, etc., to clarify the purpose and processing method of the data.

[0039] This involves the data scale, which is the total amount of data from the load device, used to measure the amount of data processing and transmission overhead.

[0040] Equipment type distinguishes the physical response characteristics of different loads (such as the chargeability and discharge characteristics of energy storage and the temperature inertia of air conditioning), data type determines whether the data needs high-frequency sampling or real-time cleaning, and data scale quantifies transmission latency and computational complexity. Clustering based on these three dimensions can ensure that load data within the same dataset has homogeneous scheduling logic and similar resource consumption curves, thereby avoiding frequent switching of computing contexts in the cloud to process heterogeneous and out-of-order data and reducing the system overhead of task scheduling.

[0041] S106, determine the target scheduling tasks corresponding to the multiple data sets respectively;

[0042] This involves target scheduling tasks, which are scheduling tasks corresponding to each data set.

[0043] By identifying target scheduling tasks corresponding to multiple data sets, the clustered data sets can be transformed into specific computational requirements that can be submitted to the cloud for execution, thereby achieving modular processing of load scheduling logic.

[0044] S108, upload multiple target scheduling tasks to the cloud for processing, and obtain the cloud scheduling parameters corresponding to each of the multiple target scheduling tasks;

[0045] For example, multiple target scheduling tasks, as well as the load data of the load devices corresponding to the edge devices of the multiple target scheduling tasks, are uploaded to the cloud for processing to obtain the cloud scheduling parameters corresponding to the multiple target scheduling tasks.

[0046] Optionally, multiple target scheduling tasks are uploaded to the cloud for processing to obtain cloud scheduling parameters corresponding to each target scheduling task. This includes: determining the uplink and downlink transmission speeds of the edge devices corresponding to each target scheduling task, where the uplink transmission speed is the data transmission speed when the corresponding edge device uploads data to the cloud, and the downlink transmission speed is the data transmission speed when the cloud sends data to the corresponding edge device; determining the cloud computing power configuration corresponding to each target scheduling task based on the task size and predicted size corresponding to each target scheduling task, as well as the uplink and downlink transmission speeds of the edge devices corresponding to each target scheduling task, where the predicted size is the predicted data size of the task processing result after the corresponding target scheduling task is processed; and processing the target scheduling tasks corresponding to each target scheduling task based on the cloud computing power configuration to obtain the cloud scheduling parameters corresponding to each target scheduling task.

[0047] This involves the cloud, which is a remote centralized computing power processing center used to receive and uniformly process the scheduling tasks uploaded by various edge devices and output the corresponding cloud scheduling parameters.

[0048] This involves cloud scheduling parameters, which are scheduling and control parameters output by the cloud after global calculation. For example, they include the optimal power curves of the load devices corresponding to each edge device (including power adjustment data at each moment within the scheduling cycle), etc., to establish a global optimization benchmark for local target scheduling parameters for edge devices, so as to achieve precise load control of cloud-edge collaboration.

[0049] This involves the uplink transmission speed, which is the data transmission speed (channel transmission rate) when the corresponding edge device uploads data to the cloud, affecting task transmission latency and control response speed.

[0050] This involves downlink transmission speed, which is the data transmission speed (channel transmission rate) when the cloud sends data to the corresponding edge device, affecting task transmission latency and control response speed.

[0051] This includes the task size, which is the amount of data and computation involved in the target scheduling task, and is used to characterize the basic resources required for task processing.

[0052] This involves the prediction scale, which is the estimated amount of data output after cloud processing, used to plan downlink transmission and storage resources in advance.

[0053] This involves cloud computing power configuration, which is the configuration of computing resources allocated by the cloud to a single target scheduling task, including the number of central processing unit (CPU) cores, the number of GPUs, the memory size, the type of computing nodes, etc., to ensure that the task is processed within the expected latency.

[0054] First, obtain the uplink and downlink transmission speeds of edge devices to accurately measure data transmission efficiency. Combine this with the task scale and predicted scale to determine the cloud computing power configuration, which enables the computing power allocation to be precisely matched with the task and transmission conditions. Because computing power is allocated on demand and transmission and computing are adapted in a coordinated manner, the waste of computing power and the latency caused by transmission blockage are avoided.

[0055] Optionally, based on the task size and predicted size corresponding to each of the multiple target scheduling tasks, and the uplink and downlink transmission speeds of the edge devices corresponding to each of the multiple target scheduling tasks, the cloud computing power configuration corresponding to each of the multiple target scheduling tasks is determined, including: calling an objective function, wherein the objective function is a function aimed at minimizing the task processing time, wherein the objective function includes an uplink transmission time item, a downlink transmission time item, and a data processing time item, wherein the uplink transmission time item is the item corresponding to the data transmission time when the corresponding edge device uploads data to the cloud, and the downlink transmission time item is the item corresponding to the data transmission time when the cloud sends data to the corresponding edge device; and based on the objective function, based on the task size and predicted size corresponding to each of the multiple target scheduling tasks, and the uplink and downlink transmission speeds of the edge devices corresponding to each of the multiple target scheduling tasks, the cloud computing power configuration corresponding to each of the multiple target scheduling tasks is determined.

[0056] This involves an objective function, which aims to minimize task processing time. The objective function includes uplink transmission time, data processing time, and downlink transmission time. Its main function is to incorporate communication latency and computation latency into a unified quantization framework, providing a clear optimization direction for solving cloud computing power configuration.

[0057] For example, the objective function can be expressed as:

[0058]

[0059] in, The task processing time corresponding to the target scheduling task; This represents the total number of edge devices corresponding to the target scheduling task. For the first The task size of the target scheduling task corresponding to each edge device; For the first Cloud computing power configuration for target scheduling tasks corresponding to each edge device; For the first The predicted scale of the target scheduling task corresponding to each edge device; For the first Uplink transmission speed of each edge device; For the first Downlink transmission speed of edge devices.

[0060] This includes an uplink transmission time term, which is the time term in the objective function that represents the time it takes for the edge device to upload task data to the cloud. This term is used to reflect the impact of uplink data transmission on the total processing time.

[0061] This includes a downlink transmission time term, which is the time term in the objective function that represents the time it takes for the cloud to send scheduling results to the edge device. This term is used to reflect the impact of downlink data transmission on the total processing time.

[0062] This includes a data processing time term, which is the time term in the objective function that represents the time taken for the cloud to perform calculations on the target scheduling task. It is used to reflect the relationship between computing power configuration and task processing efficiency.

[0063] The objective function includes three types of time terms: uplink, downlink, and data processing, fully covering the entire process of data upload, cloud computing, and instruction issuance. Therefore, it can fully represent the time consumption of the entire process. The task scale, prediction scale, and uplink and downlink transmission speed are close to the actual transmission and computing conditions. Based on its solution, the computing power configuration can be matched with the actual data transmission and computing needs, thus allowing the computing power to be accurately adapted to the actual working conditions.

[0064] Optionally, when multiple target scheduling tasks each include multiple subtasks, the objective function is expressed as:

[0065]

[0066] in, The task processing time corresponding to the target scheduling task; This represents the total number of edge devices corresponding to the target scheduling task. The total number of subtasks corresponding to the target scheduling task; For the first The first target scheduling task corresponding to the edge device The size of each subtask; For the first The first target scheduling task corresponding to the edge device Sub-computing power configuration for each sub-task; For the first Uplink transmission speed of each edge device; For the first Downlink transmission speed of individual edge devices; For the first The first target scheduling task corresponding to the edge device Sub-prediction size of each sub-task.

[0067] Optionally, the multiple subtasks of the corresponding target scheduling task correspond one-to-one with the load data in the corresponding dataset.

[0068] When the target scheduling task contains multiple subtasks, the task size, prediction size and computing power allocation of each subtask are included in the double summation term of the objective function. This can optimize the transmission and computing latency of different subtasks under the same edge device, thereby helping to solve the problems of coarse overall scheduling granularity and inaccurate computing power matching, and minimizing the overall processing time in multi-subtask scenarios.

[0069] S110, determine the matching index between each edge device and multiple target scheduling tasks, wherein the matching index is used to characterize the degree of matching between the device scheduling task corresponding to the corresponding edge device and the corresponding target scheduling task;

[0070] This involves a matching index, which is a metric that quantifies the strength of the association between edge devices and target scheduling tasks. It measures the adaptability of edge devices in executing corresponding target scheduling tasks and serves as a weight for subsequent weighted processing. For example, the matching index can be based on the degree of matching between the device scheduling task corresponding to the edge device and the corresponding target scheduling task. On this basis, it can also be further determined by comprehensively considering factors such as the computing power of the edge device, the correlation of data features, historical scheduling accuracy, transmission performance, and the type of load device.

[0071] This involves equipment scheduling tasks, which are local scheduling tasks that the edge devices themselves need to execute. For example, these equipment scheduling tasks include power allocation for the corresponding load devices, equipment start / stop control, or status detection.

[0072] By calculating the matching index between the local scheduling tasks and the target scheduling tasks of each edge device, the adaptability of each edge device to the cloud scheduling parameters corresponding to different target scheduling tasks can be quantified. This avoids the indiscriminate distribution of cloud scheduling parameters from the unified cloud to all edge devices, ensuring that subsequent scheduling parameter allocation prioritizes matching with edge devices that have higher adaptability.

[0073] S112, based on the cloud scheduling parameters corresponding to multiple target scheduling tasks and the matching index between each edge device and multiple target scheduling tasks, determine the target scheduling parameters of the load devices corresponding to each edge device, so as to schedule the corresponding load devices.

[0074] Optionally, based on the cloud scheduling parameters corresponding to multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks, the target scheduling parameters for the load devices corresponding to each edge device are determined, including: determining the reference scheduling range for the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks; determining the operation and maintenance cost and power interaction cost for the load devices corresponding to each edge device, wherein the power interaction cost is the cost incurred when the load devices corresponding to the edge devices interact with the power grid; and determining the target scheduling parameters for the load devices corresponding to each edge device based on the reference scheduling range, operation and maintenance cost, and power interaction cost for the load devices corresponding to each edge device.

[0075] This involves a reference scheduling range, which is a feasible adjustment range for each edge device after the cloud scheduling parameters are weighted by the matching index. For example, the reference scheduling range includes the upper and lower limits of power adjustment, the allowable switching range of device operating status, etc., to ensure that the final target scheduling parameters do not deviate from the global optimization direction of the cloud.

[0076] This involves operation and maintenance costs, which are the costs incurred by the load equipment under the jurisdiction of the edge device during operation and maintenance. For example, these operation and maintenance costs include equipment depreciation costs, inspection and maintenance costs, battery replacement costs, etc. Their main function is to serve as a cost item in the local optimization objective, so as to minimize equipment loss while meeting scheduling requirements.

[0077] This involves power exchange costs, which are the costs incurred when load equipment exchanges power with the power grid. For example, these power exchange costs include transmission costs, electricity consumption costs, and power dispatch costs.

[0078] By combining cloud-based scheduling parameters with matching indices to define a reference scheduling range, local control can be constrained from deviating from the global optimization direction. By introducing operation and maintenance and power interaction costs, the characteristics of equipment loss and power grid interaction can be taken into account. Since global constraints and local loss are optimized in a coordinated manner, the problem of imbalance between global and local adaptation of scheduling parameters can be solved, and the optimal target scheduling parameters that take into account both global and local factors can be obtained.

[0079] Optionally, based on the reference scheduling range, operation and maintenance cost, and power interaction cost of the load devices corresponding to each edge device, the target scheduling parameters of the load devices corresponding to each edge device are determined, including: determining the state change constraints of the load devices corresponding to each edge device; and determining the target scheduling parameters of the load devices corresponding to each edge device based on the state change constraints, reference scheduling range, operation and maintenance cost, and power interaction cost of the load devices corresponding to each edge device.

[0080] This involves state change constraints, which are physical limitations that load equipment must comply with when switching operating states or adjusting power. For example, these state change constraints include power ramp rate limits (the maximum magnitude of power change per unit time), minimum start-up and shutdown intervals for equipment, upper and lower limits of energy storage charge state, and charging and discharging depth limits, to ensure that the generated target scheduling parameters do not exceed the physical capacity of the equipment and to avoid damage to the equipment due to excessively fast or frequent adjustments.

[0081] By clearly defining the constraints on the state changes of load equipment, the physical boundaries of parameter adjustment can be limited. By combining the reference scheduling range, operation and maintenance and power interaction costs to solve the scheduling parameters, the overall control requirements, operating losses and equipment safety can be taken into account. Since the inherent physical limitations of the equipment are included, the problem of parameters exceeding the equipment's tolerance threshold can be solved, thus achieving safe load scheduling.

[0082] Through the steps S102-S112 described above, multiple data sets are formed by clustering the load data of each edge device, and a corresponding target scheduling task is matched for each set. This integrates scattered heterogeneous data into representative task units and uploads them to the cloud, thereby reducing data redundancy and computational load in cloud processing. Then, based on the scheduling parameters returned by the cloud for each task unit, and combined with the matching index between the edge device and the task, the target scheduling parameters of each edge device are allocated. Since the matching index accurately reflects the degree of fit between the device scheduling task and the target scheduling task, the global optimization parameters generated by the cloud can be adapted to each edge device in a differentiated manner. This avoids the duplication of calculations and invalid communication caused by all edge devices directly using the same cloud parameters, thus solving the technical problem of high scheduling costs in related technologies when coordinating load devices with the edge device and the cloud.

[0083] Based on the above embodiments and optional embodiments, an optional implementation method is provided, which is described in detail below.

[0084] In related technologies, scheduling load devices through cloud collaboration can achieve global unified control of a large number of load devices. However, in related technologies, when scheduling load devices based on edge devices and cloud collaboration, there is a technical problem of high scheduling costs.

[0085] For example, in related technologies, cloud-based collaborative scheduling typically uploads all load data collected by each edge device to the cloud decision center for centralized processing. Therefore, when scheduling load devices based on edge devices and cloud collaboration, there are technical problems such as large data interaction volume, heavy cloud computing load, and high scheduling costs due to uneven distribution of communication and computing resources.

[0086] There is currently no effective solution to the above problems.

[0087] In view of this, an optional embodiment of the present invention provides a load equipment scheduling method based on cloud-edge collaboration, which can effectively solve the above-mentioned technical problems.

[0088] Figure 2 This is a schematic diagram of a cloud-edge collaborative load scheduling framework in an optional embodiment of the present invention, as shown below. Figure 2 As shown, from top to bottom, the system consists of a power load management system, a cloud platform management center, edge devices, and terminal-side controllable load devices. Each level has a clear division of labor and bidirectional interaction, enabling hierarchical collaborative scheduling of controllable loads.

[0089] The cloud platform management center aims to minimize the total communication cost of controllable load demands, fully considering constraints such as the maximum number of tasks a resource node can run and the maximum number of computing resources a resource node can have, to allocate computing resources to edge computing devices. Edge-side devices (smart energy units) assess their own computing resource status; if it exceeds their own computing capacity, the edge computing device will initiate a computing resource request to the cloud platform management center. With the goal of optimizing operating costs, the center formulates optimal control instructions for each controllable load, taking into account constraints such as the operation of different types of controllable loads. Terminal-side controllable load devices include electric vehicles (EVs), air conditioners (ACLs), distributed electric heating (DEH), energy storage devices (ESS), and heat pumps (HPL), etc. These are described in detail below.

[0090] S1. Computation Task Generation and Computation Resource Request:

[0091] The equipment data of the load equipment corresponding to each edge device is determined. After the edge computing device (i.e., the edge device) summarizes the operation data (i.e., equipment data) of the multi-controllable load (i.e., load equipment) collected by the terminal layer, a computing requirement will be generated due to reasons such as monitoring of the underlying power equipment terminal and solving the multi-controllable load regulation strategy.

[0092] Edge computing devices will assess their own computing resource status. If the resource usage exceeds their capabilities, the edge computing device will request computing resources from the cloud platform management center. During this process, the data, after preprocessing, will also be uploaded to the cloud control center.

[0093] Specifically, S1 includes:

[0094] S11. Diverse controllable power loads include electric vehicles (EVs), air conditioning loads (ACLs), distributed electric heating (DEH), energy storage devices (ESS), and heat pump loads (HPL). When edge computing devices perform localized control of controllable load devices such as electric vehicles (EVs), air conditioning loads (ACLs), distributed electric heating (DEH), energy storage devices (ESS), and heat pump loads (HPL), they utilize their data analysis and processing capabilities to undertake important tasks such as data collection, data preprocessing, real-time load control, and security assurance.

[0095] S12. Edge computing devices serve as localized control centers, receiving data from various types of sensors and energy meters (i.e., the equipment data of the load devices corresponding to each edge device, such as real-time power information of multiple controllable loads like electric vehicles, air conditioning loads, distributed electric heating, energy storage devices, and heat pump load devices, as well as real-time power information of large user loads such as industrial loads).

[0096] S13. Since edge computing devices themselves have basic computing capabilities, they can perform preliminary processing and analysis on the collected data locally, identify abnormal situations or special patterns in controllable load data, and thus respond quickly, discover and resolve sudden problems that occur during the operation of multi-controllable loads.

[0097] S14. The edge computing device uploads the pre-processed multi-dimensional controllable load status data to the cloud for more advanced processing and analysis, and submits computing resource requests to the cloud platform management center (i.e., the cloud) to achieve rapid solution of edge-side control tasks.

[0098] S2, Cloud Platform Management Center issues computing resource allocation and scheduling instructions:

[0099] After receiving computing resource requests from each edge computing device, the cloud manager allocates computing resources to the edge computing devices with the goal of minimizing the total communication cost of controllable load demand tasks. This is done while fully considering constraints such as the maximum number of tasks a resource node can run and the maximum number of computing resources a resource node can have, thereby improving the efficiency of computing resource allocation. The cloud control center treats the diverse loads managed by each edge device as a whole, collaboratively participating in power system scheduling, and then issues the optimal scheduling instructions to each edge computing device.

[0100] Specifically, S2 includes:

[0101] S21. After collecting and summarizing the parallel computing tasks requested by each edge node (i.e. each edge device) at the same time, the cloud platform management center obtains the controllable load type, computing task type and computing task scale by polling or by the edge devices actively reporting.

[0102] Based on the maximum number of tasks and the maximum computing resource allocation limit of the edge nodes, the computing tasks are further divided into several categories according to the similarity of the computing tasks. That is, the equipment data of the load devices corresponding to each edge device are clustered to obtain multiple datasets.

[0103] S22. Since the tasks within each subset (i.e. each data set) after grouping are similar, a target scheduling task is determined from each data set. This target scheduling task can be the device scheduling task of the edge node with the largest data scale. The data of the edge node with the largest data scale is uploaded to the cloud platform management center for task calculation. The calculated results are then fed back to other edge nodes with the same load type and the same calculation task, thereby effectively reducing the communication and calculation overhead in the data interaction process.

[0104] Furthermore, the objective function of the cloud platform management center aims to minimize the total time overhead of computing tasks, optimize resource allocation for regrouped computing tasks, and rationally allocate computing resources. The main objective of controllable load optimization control is to minimize overall operating costs while considering internal multi-various load constraints, thereby achieving optimal load control.

[0105] For example, when multiple target scheduling tasks each include multiple subtasks, the objective function is expressed as:

[0106]

[0107] in, The task processing time (i.e., time cost) corresponding to the target scheduling task. This represents the total number of edge devices corresponding to the target scheduling task. The total number of subtasks corresponding to the target scheduling task; For the first The first target scheduling task corresponding to the edge device The size of each subtask (i.e., the data size). For the first The first target scheduling task corresponding to the edge device Sub-computing power configuration for each sub-task; For the first The uplink transmission speed (i.e., uplink channel rate) of each edge device. For the first The downlink transmission speed (i.e., downlink channel rate) of each edge device; For the first The first target scheduling task corresponding to the edge device The size of a sub-prediction for each sub-task (i.e., the amount of data returned after the task is completed).

[0108] The set of numbers for each edge-side node of a multi-element controllable power load is represented as follows: , It is the set of positive integers; Each edge node's computation task (i.e., subtask) is assigned a number, and each computation task provides the following computational resources: The goal of cloud resource optimization allocation is to cover all computing.

[0109] S3, Real-time Edge Load Control:

[0110] After receiving scheduling instructions from the cloud control center, the edge computing device formulates the optimal control instructions for each controllable load with the goal of optimizing operating costs (i.e., minimizing operating costs) and taking into account constraints such as the number of times electric vehicles can be charged and the adjustable power of air conditioners.

[0111] Through effective transmission of controllable power load data, allocation of computing resources, and real-time control of power load, the cloud-edge collaborative power load regulation platform can make full use of load data at the edge and computing resources in the cloud, thereby improving the cloud's response speed to controllable load tasks.

[0112] Specifically, S3 includes:

[0113] S31. Considering the constraints of multi-load operation, the cloud management center aims to minimize the overall operating cost to achieve optimal load control. Specifically, based on the reference scheduling range of the load equipment corresponding to each edge device, a cost optimization function is used to determine the target scheduling parameters of the load equipment corresponding to each edge device, according to the operation and maintenance cost and the power interaction cost. The cost optimization function (i.e., the optimization function for real-time optimization control of multi-controllable loads on the edge side) is expressed as follows:

[0114]

[0115] in, The total scheduling cost of the corresponding edge device is also known as the total scheduling cost on the edge side; T is the total number of time periods for load control. This represents the total number of load devices for the corresponding edge device, which is also the total number of controllable load devices within the edge area of ​​the corresponding edge device. ; For operation and maintenance costs; This refers to the cost of electricity interaction.

[0116] in, Let be the operating power of the nth load device (i.e., the controllable load device) at time t; Let be the operation and maintenance cost of the nth load device at its operating power at time t; The cost per unit power at time t; The total power consumption of all electrical loads (air conditioners, electric vehicles, electric heating, etc.) excluding energy storage devices (ESS); Let n be the charging and discharging power of the nth energy storage device (ESS) at time t; The interaction power between the edge device and the power grid at time t can be the interaction power corresponding to the scheduling command issued by the cloud control center to the edge node at time t.

[0117] in, The following formula is used to determine the value:

[0118]

[0119] in, The unit power real-time operation and maintenance coefficient of the nth load device represents the real-time loss, maintenance, energy consumption and other costs corresponding to the unit operating power; The initial cost of the nth load device is the one-time investment cost for the purchase and installation of the load device. For the entire operating lifecycle (i.e., equipment lifespan) of the nth load device; is the loss attenuation coefficient, used to characterize the annual variation of performance degradation and aging loss of load equipment during long-term use; r is the equipment type, r∈{EV,ACL,DEH,ESS,HPL}.

[0120] Among them, when The time indicates that the energy storage device is in a charging state (i.e., consuming power), and the time indicates that it is in a discharging state (i.e., outputting power). In order to track the instructions issued by the cloud, the energy storage battery adopts flexible charging and discharging measures to quantify its own operating costs. Therefore, absolute value operation is used in the cost optimization function to ensure that the cost of charging and discharging the energy storage battery is positive.

[0121] S32. Determine the state change constraints, which include constraints such as: error-free tracking constraint for scheduling instructions and the rate of change of output power of schedulable equipment.

[0122] Among them, the error-free tracking constraint of the scheduling command means that at any load control moment, the power value of the edge node (i.e., the edge device) must track the current scheduling command value issued by the cloud (i.e., conform to the reference scheduling range):

[0123]

[0124] in, The fixed load power that cannot be adjusted at time t is not involved in scheduling optimization.

[0125] The output of schedulable equipment is limited, and the controllable equipment is affected by factors such as hardware conditions and external environment. The adjustable power has definite upper and lower bounds.

[0126]

[0127] in, Let be the minimum operating power of the nth load device (i.e., the controllable load device) at time t; Let t be the maximum operating power of the nth load device (i.e., the controllable load device) at time t.

[0128] To ensure that the control equipment does not change its output power too quickly when adjusting it, in order to avoid system instability or damage to the equipment, set upper and lower limits for the allowable power change of the control equipment per unit time:

[0129]

[0130] in, This represents the minimum power adjustment range of the nth controllable load device in the rth category, i.e., the lower limit threshold of power change between adjacent scheduling times; This represents the maximum power ramp-up of the nth controllable load device in the rth category, which is the maximum allowable power change within a unit scheduling period (power ramp-up limit).

[0131] S33. Formulate real-time control instructions for each multi-dimensional load using the scheduling curve issued by the cloud control center as the control target.

[0132] The above optional implementation methods can achieve at least the following beneficial effects:

[0133] (1) Compared with related technologies, the present invention clusters the load data of each edge device into multiple data sets and matches each set with a corresponding target scheduling task. This can integrate scattered heterogeneous data into representative task units and upload them to the cloud, thereby reducing data redundancy and computational load in cloud processing. Then, based on the scheduling parameters returned by the cloud to each task unit, the target scheduling parameters of each edge device are allocated in combination with the matching index between the edge device and the task. Since the matching index accurately reflects the degree of fit between the device scheduling task and the target scheduling task of the edge device, the global optimization parameters generated by the cloud can be adapted to each edge device in a differentiated manner, avoiding the repeated calculation and invalid communication caused by all edge devices directly using the same cloud parameters. This solves the technical problem of high scheduling cost in related technologies when scheduling load devices based on the collaborative scheduling of edge devices and the cloud.

[0134] (2) Compared with related technologies, this invention can cluster the load device data corresponding to each edge device and form multiple data sets, which can organize the scattered and heterogeneous data into a unified unit according to the characteristics. Since clustering can eliminate data redundancy and disorder, it can reduce the computing and transmission burden of cloud processing, solve the problem of low efficiency caused by complex data in cloud-edge collaborative scheduling, and provide a standardized data foundation for subsequent scheduling task matching and parameter distribution.

[0135] (3) Compared with related technologies, this invention can constrain local regulation to not deviate from the global optimization direction by combining cloud scheduling parameters and matching index to define the reference scheduling range. It introduces operation and maintenance and power interaction costs, and can take into account equipment loss and power grid interaction characteristics. Since it achieves global constraint and local loss synergistic optimization, it can solve the problem of global and local adaptation imbalance of scheduling parameters, and obtain the optimal target scheduling parameters that take into account both global and local factors.

[0136] (4) Compared with related technologies, this invention can limit the physical boundary of parameter adjustment by clearly defining the state change constraints of load equipment. It can solve the scheduling parameters by combining the reference scheduling range, operation and maintenance and power interaction costs. It can take into account the global control requirements, operating losses and equipment safety. Since the inherent physical limitations of the equipment are included, the problem of parameters exceeding the equipment tolerance threshold can be solved, and safe load scheduling can be achieved.

[0137] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0138] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0139] Example 2

[0140] According to embodiments of the present invention, an apparatus for implementing the above-described cloud-edge collaborative load scheduling method is also provided. Figure 3 This is a structural block diagram of a cloud-edge collaborative load scheduling device according to an embodiment of the present invention, as shown below. Figure 3 As shown, the device includes: a first determining module 302, a second determining module 304, a third determining module 306, a fourth determining module 308, a fifth determining module 310, and a sixth determining module 312. The device will be described in detail below.

[0141] The first determining module 302 is used to determine the equipment data of the load equipment corresponding to each edge device;

[0142] The second determining module 304 is connected to the first determining module 302 and is used to cluster the device data of the load devices corresponding to each edge device to obtain multiple data sets.

[0143] The third determining module 306 is connected to the second determining module 304 and is used to determine the target scheduling tasks corresponding to the multiple data sets respectively.

[0144] The fourth determining module 308, connected to the third determining module 306, is used to upload multiple target scheduling tasks to the cloud for processing, and obtain cloud scheduling parameters corresponding to the multiple target scheduling tasks respectively.

[0145] The fifth determining module 310, connected to the fourth determining module 308, is used to determine the matching index between each edge device and multiple target scheduling tasks. The matching index is used to characterize the degree of matching between the device scheduling task corresponding to the corresponding edge device and the corresponding target scheduling task.

[0146] The sixth determining module 312, connected to the fifth determining module 310, is used to determine the target scheduling parameters of the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks, so as to schedule the corresponding load devices.

[0147] It should be noted that the first determining module 302, the second determining module 304, the third determining module 306, the fourth determining module 308, the fifth determining module 310, and the sixth determining module 312 mentioned above correspond to steps S102 to S112 in the implementation of the cloud-edge collaborative load equipment scheduling method. The instances and application scenarios implemented by multiple modules and their corresponding steps are the same, but are not limited to the content disclosed in the above embodiment 1.

[0148] Example 3

[0149] According to another aspect of the present invention, an electronic device is also provided, comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to execute instructions to implement the cloud-edge collaborative load device scheduling method of any of the above embodiments.

[0150] Example 4

[0151] According to another aspect of the present invention, a computer-readable storage medium is also provided, which, when the instructions in the computer-readable storage medium are executed by the processor of an electronic device, enables the electronic device to perform any of the above-described cloud-edge collaborative load scheduling methods.

[0152] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0153] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0154] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0155] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0156] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0157] If the integrated unit is implemented as a software functional unit 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 the present invention, in essence, or the part that contributes to the prior art, or all or part 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 described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0158] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A load scheduling method based on cloud-edge collaboration, characterized in that, include: Determine the equipment data of the load devices corresponding to each edge device; Cluster the device data of the load devices corresponding to each edge device to obtain multiple datasets; Determine the target scheduling tasks corresponding to the plurality of data sets respectively; Multiple target scheduling tasks are uploaded to the cloud for processing to obtain cloud scheduling parameters corresponding to each of the multiple target scheduling tasks. Determine the matching index between each edge device and the plurality of target scheduling tasks, wherein the matching index is used to characterize the degree of matching between the device scheduling task corresponding to the corresponding edge device and the corresponding target scheduling task; Based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks, and the matching index between each edge device and the multiple target scheduling tasks, the target scheduling parameters of the load devices corresponding to each edge device are determined for scheduling the corresponding load devices.

2. The method according to claim 1, characterized in that, The process of uploading multiple target scheduling tasks to the cloud for processing, and obtaining cloud scheduling parameters corresponding to each of the multiple target scheduling tasks, includes: The uplink and downlink transmission speeds of the edge devices corresponding to the multiple target scheduling tasks are determined respectively, wherein the uplink transmission speed is the data transmission speed when the corresponding edge device uploads data to the cloud, and the downlink transmission speed is the data transmission speed when the cloud sends data to the corresponding edge device; Based on the task size and predicted size corresponding to the multiple target scheduling tasks, as well as the uplink and downlink transmission speeds of the edge devices corresponding to the multiple target scheduling tasks, the cloud computing power configuration corresponding to the multiple target scheduling tasks is determined, wherein the predicted size is the predicted data size of the task processing result after the corresponding target scheduling task is processed. Based on the cloud computing power configuration corresponding to the multiple target scheduling tasks, the target scheduling tasks corresponding to the multiple target scheduling tasks are processed to obtain the cloud scheduling parameters corresponding to the multiple target scheduling tasks.

3. The method according to claim 2, characterized in that, The step of determining the cloud computing power configuration corresponding to each of the multiple target scheduling tasks based on the task size and predicted size corresponding to each of the multiple target scheduling tasks, and the uplink and downlink transmission speeds of the edge devices corresponding to each of the multiple target scheduling tasks, includes: The objective function is invoked, wherein the objective function is a function aimed at minimizing the task processing time. The objective function includes an uplink transmission time item, a downlink transmission time item, and a data processing time item. The uplink transmission time item is the item corresponding to the data transmission time when the corresponding edge device uploads data to the cloud, and the downlink transmission time item is the item corresponding to the data transmission time when the cloud sends data to the corresponding edge device. Based on the objective function, the cloud computing power configuration corresponding to each of the multiple target scheduling tasks is determined based on the task size and predicted size corresponding to each of the multiple target scheduling tasks, as well as the uplink and downlink transmission speeds of the edge devices corresponding to each of the multiple target scheduling tasks.

4. The method according to claim 3, characterized in that, When the multiple target scheduling tasks each include multiple subtasks, the objective function is expressed as: in, The task processing time corresponding to the target scheduling task; This represents the total number of edge devices corresponding to the target scheduling task. The total number of subtasks corresponding to the target scheduling task; For the first The first target scheduling task corresponding to the edge device The size of each subtask; For the first The first target scheduling task corresponding to the edge device Sub-computing power configuration for each sub-task; For the first Uplink transmission speed of each edge device; For the first Downlink transmission speed of individual edge devices; For the first The first target scheduling task corresponding to the edge device Sub-prediction size of each sub-task.

5. The method according to claim 1, characterized in that, The step of determining the target scheduling parameters for the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks includes: Based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks, and the matching index between each edge device and the multiple target scheduling tasks, the reference scheduling range of the load device corresponding to each edge device is determined. Determine the operation and maintenance cost and power interaction cost of the load equipment corresponding to each edge device, wherein the power interaction cost is the cost incurred when the load equipment corresponding to the edge device interacts with the power grid; Based on the reference scheduling range, operation and maintenance cost, and power interaction cost of the load devices corresponding to each edge device, the target scheduling parameters of the load devices corresponding to each edge device are determined.

6. The method according to claim 5, characterized in that, The determination of target scheduling parameters for the load devices corresponding to each edge device, based on the reference scheduling range, operation and maintenance costs, and power interaction costs of the load devices corresponding to each edge device, includes: Determine the state change constraints of the load devices corresponding to each edge device; Based on the state change constraints of the load devices corresponding to each edge device, and taking into account the scheduling range, operation and maintenance costs, and power interaction costs, the target scheduling parameters of the load devices corresponding to each edge device are determined.

7. The method according to any one of claims 1 to 6, characterized in that, The clustering of device data corresponding to each edge device yields multiple data sets, including: Determine the device type of the load device corresponding to each edge device, as well as the data type and data size of the device data of the load device corresponding to each edge device; Based on the device type of the load device corresponding to each edge device, and the data type and data scale of the device data of the load device corresponding to each edge device, the device data of the load device corresponding to each edge device are clustered to obtain multiple data sets.

8. A load scheduling device based on cloud-edge collaboration, characterized in that, include: The first determining module is used to determine the equipment data of the load equipment corresponding to each edge device; The second determining module is used to cluster the device data of the load devices corresponding to each edge device to obtain multiple data sets; The third determining module is used to determine the target scheduling tasks corresponding to the plurality of data sets respectively; The fourth determining module is used to upload multiple target scheduling tasks to the cloud for processing, and obtain cloud scheduling parameters corresponding to the multiple target scheduling tasks respectively; The fifth determining module is used to determine the matching index between each edge device and the plurality of target scheduling tasks, wherein the matching index is used to characterize the degree of matching between the device scheduling task corresponding to the corresponding edge device and the corresponding target scheduling task; The sixth determining module is used to determine the target scheduling parameters of the load devices corresponding to each edge device based on the cloud scheduling parameters corresponding to the multiple target scheduling tasks and the matching index between each edge device and the multiple target scheduling tasks, so as to schedule the corresponding load devices.

9. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the cloud-edge collaborative load scheduling method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the cloud-edge collaborative load scheduling method as described in any one of claims 1 to 7.

Images