Task scheduling strategy and system in cloud-edge collaborative centralized control system

Abstract

The present invention provides a task scheduling strategy and system in a cloud-edge collaborative centralized control system, which belongs to the field of cloud-edge collaborative smart grid sites in the Internet of Things, and solves the problems of resource allocation and task processing timeliness in the dynamically changing cloud-edge system centralized control system. The main method adopted is as follows: by calculating the transmission rate, transmission delay, computing delay, terminal processing energy consumption, and task processing completion delay of the unloaded task, the objective function and constraints are finally determined, and the particle swarm algorithm is used to solve the objective function and output the optimal solution. The present invention determines a set of optimal parameter configurations, including the optimal transmission power, network bandwidth allocation ratio, and computing resource allocation ratio, by applying an optimized particle swarm algorithm, so as to minimize the total delay of task processing within the framework of the cloud-edge collaborative centralized control system.

Description

Technical Field

[0001] The present invention relates to the field of IoT cloud-edge collaborative smart grid sites, and specifically to a task scheduling strategy and system in a cloud-edge collaborative centralized control system. Background Art

[0002] In the context of the global energy system transformation and the deep integration of digital technology, the construction of smart grids has become the cornerstone of the modernization of power systems. Smart grid technology has greatly enhanced the automation, interactivity and flexibility of power grids by deeply integrating power and information flows. Among them, the cloud-edge collaborative centralized control system, as the key to innovation in operation and maintenance models, integrates the powerful computing power of cloud computing and the instant response capability of edge computing, realizes the immediate collection, processing, analysis and decision-making of power grid operation data, and significantly improves operation and maintenance efficiency and system reliability.

[0003] However, the expansion of the scale of smart grids and the increase in structural complexity have brought new challenges to the cloud-edge collaborative centralized control system. In particular, under the complex distributed architecture, the effective scheduling of smart inspection tasks and the balanced allocation of resources have become technical difficulties. Smart inspection covers a wide range of operation and maintenance activities, from equipment health monitoring to fault prediction and even pre-maintenance. It is data-intensive, has diverse processing requirements, and has strict real-time requirements. The traditional static scheduling mechanism is difficult to cope with this dynamically changing operation and maintenance environment, which can easily lead to problems such as uneven resource allocation and increased processing delays, which seriously threaten the safe and stable operation of the power grid.

[0004] In addition, ensuring efficient and stable transmission of the in-station communication network is another core element of intelligent operation and maintenance. The massive data generated by intelligent inspection activities needs to flow quickly between the cloud, edge, and terminals. Any communication delay or interruption may weaken the timeliness of decision-making, especially in complex electromagnetic environments, network congestion, or hardware failure scenarios. Maintaining communication quality has become a technical obstacle that needs to be overcome.

[0005] In view of this, exploring a mechanism that can adapt to large-scale distributed architecture, realize efficient scheduling and load balancing of tasks, and ensure high reliability and low latency of intra-station communication has become a core issue in smart grid technology research. Summary of the invention

[0006] The technical problem to be solved by the present invention is how to ensure uniform resource allocation and timely processing in a dynamically changing cloud-edge collaborative centralized control system.

[0007] The present invention solves the above technical problems through the following technical means:

[0008] The task scheduling strategy in the cloud-edge collaborative centralized control system includes the following steps:

[0009] Step 1. Offload task transfer rate calculation:

[0010] Assume that the mth (m∈M,M={1,2,...,M}) device within the communication range of the nth (n∈N,N={1,2,...,N})th grid station calculates the task Z that needs to be unloaded n,m The uplink transmission rate R sent to the edge server in the power grid site n,m ;

[0011] Step 2. Calculate the transmission delay: This includes calculating the wireless transmission delay of the terminal device transmitting the task to the edge server in the smart site. The wired transmission delay of the task uploaded from the intelligent site to the centralized control master station through wired communication

[0012]

[0013] Step 3. Calculate the computational latency: This includes the computational latency of the task being processed by the edge server in the smart site. The computational delay of the task being processed by the centralized control master server

[0014] Step 4. Calculate the terminal processing energy consumption: Calculate the energy consumption E of the task data transmission operation initiated by the terminal device to the smart site n,m ;

[0015] Step 5. Calculate the task processing completion delay and determine the objective function and constraints:

[0016] The offloading strategy of each task is determined by x n,m ∈{0,1} means that when x n,m = 0, the task is offloaded to the nth intelligent station for processing, and the task processing completion delay is T n,m for:

[0017]

[0018] When x n,m =1, the task is unloaded to the centralized control master station for processing, and the task processing completion delay is T c,n,m for:

[0019]

[0020] Among them, the task calculation density of the terminal device is defined as K, f n,m f is the CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site; c,n,m The CPU frequency assigned by the centralized control master station server to the mth terminal device task within the communication range of the nth smart site;

[0021] The processing completion delay of a single task for:

[0022]

[0023] System delay T total The largest processing delay among all tasks of the terminal device:

[0024]

[0025] The objective function of minimizing system delay is:

[0026]

[0027] The constraints of the objective function are:

[0028]

[0029] Where C1 represents the power constraint, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The maximum transmission power; C2 represents the bandwidth constraint, that is, the sum of the bandwidths of each channel is less than the total bandwidth of the system, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, where B max is the total system bandwidth; C3 represents the energy consumption constraint, where is the maximum energy consumption of each terminal device; C4 represents the computing resource constraint of the smart site edge server, where is the maximum computing resource of the smart site edge server; C5 represents the computing resource constraint of the centralized control master station server, where: is the maximum computing resource of the centralized control master station server; C6 represents the processing delay constraint of each task, where τ n,m is the delay constraint of the mth terminal device task within the communication range of the nth smart station;

[0030] Step 6. Use the improved particle swarm algorithm to solve the objective function and output the optimal transmission power, network bandwidth allocation ratio, and computing resource allocation ratio.

[0031] Furthermore, the transmission rate R in step 1 n,m The calculation formula is as follows:

[0032]

[0033] In the formula, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, p n,mAssign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, h n,m is the uplink channel gain, v is the path loss exponent, N0 is the channel Gaussian white noise, d n,m is the communication distance between the nth smart station and the mth terminal device within its communication range.

[0034] Furthermore, in step 2, the wireless transmission delay The calculation formula is as follows:

[0035]

[0036] Wired transmission delay The calculation formula is as follows:

[0037]

[0038] In the formula, R c,n,m is the propagation speed of electromagnetic waves.

[0039] Furthermore, in step 3, the process of calculating the delay is as follows:

[0040] First, define the task computing density of the terminal device as K, which is the number of CPU cycles required to process one bit.

[0041] The computational latency of tasks being processed by edge servers within smart sites for:

[0042]

[0043] In the formula, f n,m The CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site;

[0044] The computational delay of the task being processed by the centralized control master server for:

[0045]

[0046] In the formula, f c,n,m The CPU frequency assigned by the centralized control master server to the mth terminal device task within the communication range of the nth smart site.

[0047] Furthermore, in step 4, the energy consumption E n,m The calculation formula is:

[0048]

[0049] The present invention also provides a task scheduling system in a cloud-edge collaborative centralized control system, including:

[0050] Offload task transfer rate calculation module:

[0051] Assume that the mth (m∈M,M={1,2,...,M}) device within the communication range of the nth (n∈N,N={1,2,...,N})th grid station calculates the task Z that needs to be unloaded n,m The uplink transmission rate R sent to the edge server in the power grid site n,m ;

[0052] Calculation of transmission delay module: including calculation of wireless transmission delay of terminal equipment transmitting tasks to edge servers in smart sites The wired transmission delay of the task uploaded from the intelligent site to the centralized control master station through wired communication

[0053]

[0054] Module for calculating computational delay: including the computational delay of tasks processed by edge servers in smart sites The computational delay of the task being processed by the centralized control master server

[0055] Calculation terminal processing energy consumption module: Calculate the energy consumption E of the task data transmission operation initiated by the terminal device to the smart site n,m ;

[0056] Objective function and constraint calculation module, calculates the task processing completion delay and determines the objective function and constraint conditions:

[0057] The offloading strategy of each task is determined by x n,m ∈{0,1} means that when x n,m = 0, the task is offloaded to the nth intelligent station for processing, and the task processing completion delay is T n,m for:

[0058]

[0059] When x n,m =1, the task is unloaded to the centralized control master station for processing, and the task processing completion delay is T c,n,m for:

[0060]

[0061] Among them, the task calculation density of the terminal device is defined as K, f n,m f is the CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site; c,n,m The CPU frequency assigned by the centralized control master station server to the mth terminal device task within the communication range of the nth smart site;

[0062] The processing completion delay of a single task for:

[0063]

[0064] System delay T total The largest processing delay among all tasks of the terminal device:

[0065]

[0066] The objective function of minimizing system delay is:

[0067]

[0068] The constraints of the objective function are:

[0069]

[0070] Where C1 represents the power constraint, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The maximum transmission power; C2 represents the bandwidth constraint, that is, the sum of the bandwidths of each channel is less than the total bandwidth of the system, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, where B max is the total system bandwidth; C3 represents the energy consumption constraint, where is the maximum energy consumption of each terminal device; C4 represents the computing resource constraint of the smart site edge server, where is the maximum computing resource of the smart site edge server; C5 represents the computing resource constraint of the centralized control master station server, where: is the maximum computing resource of the centralized control master station server; C6 represents the processing delay constraint of each task, where τ n,m is the delay constraint of the mth terminal device task within the communication range of the nth smart station;

[0071] The optimization module is used to solve the objective function using the improved particle swarm algorithm and output the optimal transmission power, network bandwidth allocation ratio and computing resource allocation ratio.

[0072] Furthermore, the transmission rate R in the offload task transmission rate calculation module n,m The calculation formula is as follows:

[0073]

[0074] In the formula, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, h n,m is the uplink channel gain, v is the path loss exponent, N0 is the channel Gaussian white noise, d n,m is the communication distance between the nth smart station and the mth terminal device within its communication range.

[0075] Furthermore, in the transmission delay calculation module, the wireless transmission delay The calculation formula is as follows:

[0076]

[0077] Wired transmission delay The calculation formula is as follows:

[0078]

[0079] In the formula, R c,n,m is the propagation speed of electromagnetic waves.

[0080] Furthermore, in the module for calculating the calculation delay, the process of calculating the calculation delay is as follows:

[0081] First, define the task computing density of the terminal device as K, which is the number of CPU cycles required to process one bit.

[0082] The computational latency of tasks being processed by edge servers within smart sites for:

[0083]

[0084] In the formula, f n,m The CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site;

[0085] The computational delay of the task being processed by the centralized control master server for:

[0086]

[0087] In the formula, f c,n,m The CPU frequency assigned by the centralized control master server to the mth terminal device task within the communication range of the nth smart site.

[0088] Furthermore, in the energy consumption module of the computing terminal processing, the energy consumption E n,m The calculation formula is:

[0089]

[0090] The advantages of the present invention are:

[0091] The present invention determines a set of optimal parameter configurations, including the optimal transmission power, network bandwidth allocation ratio, and computing resource allocation ratio, by applying an optimized particle swarm algorithm, so as to minimize the total delay of task processing within the framework of a cloud-edge collaborative centralized control system. This strategy not only promotes the maximum efficiency of resource utilization, but also ensures that the system can respond quickly to various terminal business needs, thereby maintaining excellent operating performance and agility in a complex service environment. The breakthrough progress of this technology will not only greatly enhance the automation and intelligence level of power grid operation and maintenance, but will also effectively reduce operation and maintenance costs, enhance the resilience and sustainability of the power grid, and provide strong support for building a safer, environmentally friendly, and efficient energy Internet. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 This is an architecture diagram of a cloud-edge collaborative centralized control system in an embodiment of the present invention;

[0093] Figure 2 This is a logic diagram for executing the scheduling strategy in an embodiment of the present invention. DETAILED DESCRIPTION

[0094] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0095] This embodiment provides a task scheduling strategy based on an adaptive particle swarm algorithm in a cloud-edge collaborative centralized control system. The architecture of the cloud-edge collaborative centralized control system is as follows: Figure 1 As shown:

[0096] The cloud-edge collaborative centralized control system is a highly integrated smart grid management architecture that achieves efficient collaborative operations from the cloud to the terminal through multi-level resource allocation and information interaction. The system includes a cloud center control layer, N edge computing layers, and N·M terminal device layers.

[0097] The cloud center control layer, namely the centralized control station and the intelligent patrol station, focuses on global data integration, storage, analysis, strategy planning, task scheduling and resource optimization. Relying on advanced cloud computing capabilities, this layer can efficiently process massive amounts of data, conduct in-depth analysis, and lay a solid foundation for intelligent decision-making. Among them, the intelligent patrol station focuses on intelligent inspection tasks and data analysis to ensure full-time, all-round monitoring and immediate response of the substation environment, improving the intelligence and emergency efficiency of operation and maintenance.

[0098] The edge computing layer covers intelligent sites such as core substations, converter stations, and auxiliary control hosts. The following is an introduction to a core substation as an example. The core substation is specifically designed for sites with personnel on duty and direct intervention by the operation and maintenance team. It integrates an advanced intelligent inspection host system and a series of peripheral terminal devices, including but not limited to intelligent monitoring vision systems, ubiquitous sensor networks, etc. These configurations are close to the data generation point, perform data processing and analysis, and realize real-time monitoring and rapid response of key substation equipment, thereby alleviating the pressure of cloud computing and accelerating the decision-making cycle. The auxiliary control host of the substation is responsible for the acquisition of basic information, local operation control and data processing. As a bridge for communication between the cloud and terminal devices, the auxiliary control host system ensures the efficiency and security of data flow, and provides the necessary technical support and guarantee for autonomous operation and management at the site level.

[0099] The terminal equipment layer is directly located at the substation site, including but not limited to transformers, circuit breakers, relays, smart meters, and a variety of monitoring and control equipment. These devices establish a communication link with the edge computing layer, upload operating status, energy consumption data, and abnormal alarms in real time, and receive control instructions from superiors. The present invention assumes that these terminal devices use frequency division multiplexing access technology to communicate with the edge computing layer, allocate independent frequency bandwidths for different signals, and ensure that the signals of each channel do not interfere with each other.

[0100] When the terminal equipment in the substation faces data processing tasks during operation, it is necessary to implement a task offloading strategy, that is, to migrate the business to the edge computing node or cloud server for processing. This process is based on multi-dimensional factors such as business type, data size, business delay constraints, bandwidth, and computing resources of each server to determine the most suitable business processing location, whether to offload to the edge side for local processing or to the cloud for processing. When multiple terminal devices unload services in parallel, the system faces the dual challenges of efficient configuration of network bandwidth and optimal allocation of server computing resources. At the same time, this embodiment takes into account energy consumption constraints and introduces a transmission power control mechanism, aiming to coordinate resource management in multiple dimensions through optimization models. Therefore, an improved adaptive particle swarm optimization algorithm is used to calculate the optimal power control, bandwidth allocation ratio, and computing resource allocation ratio to minimize the total delay of business processing, achieve reasonable resource allocation, and thus improve the performance of the entire system.

[0101] like Figure 2 As shown in the figure, a task scheduling strategy based on adaptive particle swarm algorithm in cloud-edge collaborative centralized control system is implemented in the following steps:

[0102] Step 1: Transfer Rate

[0103] According to Shannon's theorem, the mth (m∈M, M={1,2,...,M})th device within the communication range of the nth (n∈N, N={1,2,...,N})th substation calculates the task Z that needs to be unloaded n,m The uplink transmission rate R sent to the edge server in the substation n,m for:

[0104]

[0105] In the formula, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth substation, p n,m Assign task Z to the mth terminal device within the communication range of the nth substation n,m The transmission power, h n,m is the uplink channel gain, v is the path loss exponent, N0 is the channel Gaussian white noise, d n,m is the communication distance between the nth substation and the mth terminal device within its communication range.

[0106] Step 2: Transmission Delay

[0107] Transmission delay of the terminal device transmitting the task to the edge server in the substation for:

[0108]

[0109] The task is uploaded from the substation to the centralized control master station through wired communication, and the transmission delay for:

[0110]

[0111] In the formula, R c,n,m The propagation speed of electromagnetic waves is approximately equal to the speed of light, which is 3.0×10 8 (m / s).

[0112] Step 3: Calculate latency

[0113] First, the task computing density of the terminal device is defined as K, which is the number of CPU cycles required to process one bit.

[0114] The computational latency of the task being processed by the edge server in the substation for:

[0115]

[0116] In the formula, f n,m The CPU frequency allocated to the mth terminal device task by the edge server in the nth substation.

[0117] The computational delay of the task being processed by the centralized control master server for:

[0118]

[0119] In the formula, f c,n,m The CPU frequency assigned by the centralized control master station server to the mth terminal device task within the communication range of the nth substation.

[0120] Step 4: Terminal processing energy consumption

[0121] During the task scheduling process, the terminal device initiates the transmission of task data to the substation, and the energy consumption associated with this link is E n,m for:

[0122]

[0123] Step 5: Delay in processing each task

[0124] Given that the amount of task data after processing is significantly reduced, compared to the original task size, when considering system performance indicators, it can be reasonably inferred that the delay in feedback of the results after the task is processed has little impact on the overall delay and should be ignored so that we can focus on other more critical delay components. Therefore, the delay for completing each task includes the task upload and task calculation delays.

[0125] The offloading strategy of each task is determined by x n,m ∈{0,1} means that when x n,m = 0, the task is unloaded to the nth substation for processing, and the task processing completion delay is T n,m for:

[0126]

[0127] When x n,m =1, the task is unloaded to the centralized control master station for processing, and the task processing completion delay is T c,n,m for:

[0128]

[0129] Step 5: Determine the objective function and constraints

[0130] Given the heterogeneity of the attributes of each task and its unique latency requirements, when a task is offloaded to a substation for processing but cannot meet the established latency constraints, the task needs to be strategically transferred to the central control master station to ensure that the strict latency requirements are met. for:

[0131]

[0132] Since multiple terminal devices process tasks in parallel, the system delay is T total It is the largest task processing delay among all terminal devices:

[0133]

[0134] This embodiment jointly optimizes the transmission power of the terminal device, the network bandwidth allocation, and the server computing resource allocation. The system delay minimization problem of the strategy of this embodiment is expressed as follows:

[0135]

[0136] Where C1 represents the power constraint, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The maximum transmission power; C2 represents the bandwidth constraint, that is, the sum of the bandwidths of each channel is less than the total bandwidth of the system, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, where B max is the total system bandwidth; C3 represents the energy consumption constraint, where is the maximum energy consumption of each terminal device; C4 represents the computing resource constraint of the smart site edge server, where is the maximum computing resource of the smart site edge server; C5 represents the computing resource constraint of the centralized control master station server, where: is the maximum computing resource of the centralized control master station server; C6 represents the processing delay constraint of each task, where τ n,m is the delay constraint of the mth terminal device task within the communication range of the nth smart station;

[0137] Step 6: Use improved particle swarm algorithm to solve

[0138] Under the given constraint framework, this embodiment uses an improved adaptive particle swarm optimization algorithm to solve the objective function, aiming to discover the optimal transmission power configuration, bandwidth division ratio and computing resource allocation scheme. Compared with the traditional particle swarm optimization algorithm, this embodiment integrates an adaptive mechanism to achieve dynamic control of inertia weight, thereby enhancing the flexibility and efficiency of the algorithm. The specific steps are as follows:

[0139] (1) Initialization

[0140] Initialize the particle position and velocity, set the algorithm parameters, including the inertia weight ω i , cognitive weight c1, social weight c2.

[0141] (2) Calculate the fitness value

[0142] Use the objective function shown in formula 11 in step 5 above to calculate the fitness value T of each particle i .

[0143] (3) Optimal Update

[0144] Update the individual optimal value p of each particle best,i and the global optimal value g best .

[0145] (4) Improved adaptive inertia weight

[0146] The inertia weight is dynamically adjusted according to the fitness distribution of the current particle swarm. Inertia weight ω i The specific expression is as follows:

[0147]

[0148] In the formula, ω min and ω max Respectively represent the minimum and maximum values ​​of the inertia weight, θ is a random number in the interval [0,2], is the average fitness of all particles, using the function The fitness value is converted into a weight adjustment factor to ensure that particles with high fitness within a certain range, that is, particles that exceed the average fitness of the group, are given a smaller weight to promote local search; conversely, particles with low fitness are assigned a larger weight to encourage global exploration behavior.

[0149] (5) Update speed and position

[0150] Adaptive inertia weight is used to update the velocity and position of particles. Accordingly, the particle velocity update formula can be adjusted as follows:

[0151]

[0152] In the formula, represents the speed of the ath iteration of the i-th particle, c1 and c2 represent cognitive weight and social weight respectively, r1 and r2 represent the random numbers in the [0,1] interval generated in each iteration respectively, and the individual optimal value is p best,i , the global optimal value is g best , represents the position of the ith particle at the ath iteration.

[0153] (6) Iteration

[0154] Repeat the above steps (2) to (5) until the termination condition is met.

[0155] (7) Result output

[0156] Output the global optimal particle position, that is, the optimal solution.

[0157] Step 7: Get the optimization result

[0158] By applying the optimized particle swarm algorithm, a set of optimal parameter configurations are determined, including the optimal transmission power, network bandwidth allocation ratio, and computing resource allocation ratio, so as to minimize the total latency of task processing within the framework of the cloud-edge collaborative centralized control system. This strategy not only promotes the maximum efficiency of resource utilization, but also ensures that the system can quickly respond to various terminal business needs, thereby maintaining excellent operational performance and agility in a complex service environment.

[0159] The above embodiments are only used to illustrate the technical solution of the present embodiment, rather than to limit it. Although the present embodiment is described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. Task scheduling strategy in cloud-edge collaborative centralized control system, It is characterized in that The following steps are involved: Step 1. Offload task transfer rate calculation: Assume that the mth device within the communication range of the nth power grid site, n∈N, N={1,2,...,N}, m∈M, M={1,2,...,M}; Calculate the tasks Z that will need to be offloaded n,m The uplink transmission rate R sent to the edge server in the power grid site n，m ; Step 2. Calculate the transmission delay: This includes calculating the wireless transmission delay of the terminal device transmitting the task to the edge server in the smart site. The wired transmission delay of the task uploaded from the intelligent site to the centralized control master station through wired communication Step 3. Calculate the computational latency: This includes the computational latency of the task being processed by the edge server in the smart site. The computational delay of the task being processed by the centralized control master server Step 4. Calculate the terminal processing energy consumption: Calculate the energy consumption E of the task data transmission operation initiated by the terminal device to the smart site n,m ; Step 5. Calculate the task processing completion delay and determine the objective function and constraints: The offloading strategy of each task is determined by x n,m ∈{0,1} means that when x n,m = 0, the task is offloaded to the nth intelligent station for processing, and the task processing completion delay is T n,m for: When x n,m =1, the task is unloaded to the centralized control master station for processing, and the task processing completion delay is T c,n,m for: Among them, the task calculation density of the terminal device is defined as K, f n,m f is the CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site; c,n,m is the CPU frequency assigned by the centralized control master server to the mth terminal device task within the communication range of the nth intelligent site, R c,n,m is the propagation speed of electromagnetic waves; The processing completion delay of a single task for: System delay T total The largest processing delay among all tasks of the terminal device: The objective function of minimizing system delay is: The constraints of the objective function are: In the formula, C 1 represents the power constraint, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The maximum transmission power of C 2 represents bandwidth constraint, that is, the sum of the bandwidths of each channel is less than the total bandwidth of the system, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, where B max is the total system bandwidth; C 3 represents the energy consumption constraint, where is the maximum energy consumption of each terminal device; C 4 represents the computing resource constraints of the smart site edge server, where f n max is the maximum computing resource of the smart site edge server; C 5 represents the computing resource constraints of the centralized control master station server, where f c max is the maximum computing resource of the centralized control master station server; C 6 represents the processing delay constraint of each task, where τ n,m is the delay constraint of the mth terminal device task within the communication range of the nth smart station; Step 6. Use the improved particle swarm algorithm to solve the objective function and output the optimal transmission power, network bandwidth allocation ratio, and computing resource allocation ratio.

2. According to the task scheduling strategy in the cloud-edge collaborative centralized control system of claim 1, It is characterized in that The transmission rate R in step 1 n,m The calculation formula is as follows: In the formula, h n,m is the uplink channel gain, v is the path loss exponent, N 0 is the channel Gaussian white noise, d n,m is the communication distance between the nth smart station and the mth terminal device within its communication range.

3. According to the task scheduling strategy in the cloud-edge collaborative centralized control system of claim 2, It is characterized in that In step 2, the wireless transmission delay The calculation formula is as follows: Wired transmission delay The calculation formula is as follows:

4. According to the task scheduling strategy in the cloud-edge collaborative centralized control system of claim 3, It is characterized in that In step 3, the process of calculating the delay is as follows: First, the task computing density of the terminal device is defined as K, which is the number of CPU cycles required to process one bit. The computational latency of tasks being processed by edge servers within smart sites for: In the formula, f n,m The CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site; The computational delay of the task being processed by the centralized control master server for: In the formula, f c,n,m The CPU frequency assigned by the centralized control master server to the mth terminal device task within the communication range of the nth smart site.

5. According to the task scheduling strategy in the cloud-edge collaborative centralized control system of claim 4, It is characterized in that In step 4, the energy consumption E n,m The calculation formula is:

6. Task scheduling system in cloud-edge collaborative centralized control system, It is characterized in that include: Offload task transfer rate calculation module: Assume that the mth device within the communication range of the nth power grid site, n∈N, N={1,2,...,N}, m∈M, M={1,2,...,M}; Calculate the tasks Z that will need to be offloaded n,m The uplink transmission rate R sent to the edge server in the power grid site n，m ; Calculation of transmission delay module: including calculation of wireless transmission delay of terminal equipment transmitting tasks to edge servers in smart sites The wired transmission delay of the task uploaded from the intelligent site to the centralized control master station through wired communication Module for calculating computational delay: including the computational delay of tasks processed by edge servers in smart sites The computational delay of the task being processed by the centralized control master server Calculation terminal processing energy consumption module: Calculate the energy consumption E of the task data transmission operation initiated by the terminal device to the smart site n,m ; Objective function and constraint calculation module, calculates the task processing completion delay and determines the objective function and constraint conditions: The offloading strategy of each task is determined by x n,m ∈{0,1} means that when x n,m = 0, the task is offloaded to the nth intelligent station for processing, and the task processing completion delay is T n,m for: When x n,m =1, the task is unloaded to the centralized control master station for processing, and the task processing completion delay is T c,n,m for: Among them, the task calculation density of the terminal device is defined as K, f n,m f is the CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site; c,n,m is the CPU frequency assigned by the centralized control master server to the mth terminal device task within the communication range of the nth intelligent site, R c,n,m is the propagation speed of electromagnetic waves; The processing completion delay of a single task for: System delay T total The largest processing delay among all tasks of the terminal device: The objective function of minimizing system delay is: The constraints of the objective function are: In the formula, C 1 represents the power constraint, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The maximum transmission power of C 2 represents bandwidth constraint, that is, the sum of the bandwidths of each channel is less than the total bandwidth of the system, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, where B max is the total system bandwidth; C 3 represents the energy consumption constraint, where is the maximum energy consumption of each terminal device; C 4 represents the computing resource constraints of the smart site edge server, where f n max is the maximum computing resource of the smart site edge server; C 5 represents the computing resource constraints of the centralized control master station server, where f c max is the maximum computing resource of the centralized control master station server; C 6 represents the processing delay constraint of each task, where τ n,m is the delay constraint of the mth terminal device task within the communication range of the nth smart station; The optimization module is used to solve the objective function using the improved particle swarm algorithm and output the optimal transmission power, network bandwidth allocation ratio and computing resource allocation ratio.

7. The task scheduling system in the cloud-edge collaborative centralized control system according to claim 6, It is characterized in that The transmission rate R in the offload task transmission rate calculation module n,m The calculation formula is as follows: In the formula, B n,m is the bandwidth allocated to the mth terminal device within the communication range of the nth smart station, p n,m Assign task Z to the mth terminal device within the communication range of the nth smart station n,m The transmission power, h n,m is the uplink channel gain, v is the path loss exponent, N 0 is the channel Gaussian white noise, d n,m is the communication distance between the nth smart station and the mth terminal device within its communication range.

8. The task scheduling system in the cloud-edge collaborative centralized control system according to claim 7, It is characterized in that In the transmission delay calculation module, the wireless transmission delay The calculation formula is as follows: Wired transmission delay The calculation formula is as follows:

9. The task scheduling system in the cloud-edge collaborative centralized control system according to claim 8, It is characterized in that In the calculation delay calculation module, the calculation delay calculation process is as follows: First, the task computing density of the terminal device is defined as K, which is the number of CPU cycles required to process one bit. The computational latency of tasks being processed by edge servers within smart sites for: In the formula, f n,m The CPU frequency assigned to the mth terminal device task by the edge server in the nth smart site; The computational delay of the task being processed by the centralized control master server for: In the formula, f c,n,m The CPU frequency assigned by the centralized control master server to the mth terminal device task within the communication range of the nth smart site.

10. The task scheduling system in the cloud-edge collaborative centralized control system according to claim 9, It is characterized in that In the energy consumption module of the computing terminal processing, the energy consumption E n,m The calculation formula is:

Images