Spinning data acquisition time delay optimization method and system

Abstract

The invention provides a spinning data acquisition time delay optimization method and system, and relates to the technical field of data acquisition optimization, and the method comprises the steps: collecting a plurality of groups of spinning data, and taking each group of spinning data as a task unit; edge end time delay of each task unit during edge end processing and cloud end time delay of each task unit during cloud end processing are determined; according to the edge end time delay and the cloud end time delay, determining the total time delay of each task unit during processing; constructing a spinning data acquisition time delay optimization model; setting constraint conditions of the spinning data acquisition time delay optimization model; optimizing the spinning data acquisition time delay optimization model through a PSO-SA algorithm by taking minimization of the total time delay as a target under the constraint of the constraint condition to obtain an optimal task allocation scheme and an optimal computing resource allocation scheme; and performing spinning data acquisition according to the optimal task allocation scheme and the optimal computing resource allocation scheme.

Description

Technical Field

[0001] The present invention relates to the technical field of data collection optimization, and in particular to a spinning data collection delay optimization method and system. Background Art

[0002] With the rapid development of the textile industry, especially the accelerated advancement of intelligent transformation, many companies have adopted full-process information management to improve operational efficiency and market responsiveness. Spinning, as an important part of textile production, constitutes the foundation of the entire industrial chain. The promotion and application of informatization in the spinning field not only improves production efficiency, but also provides companies with significant advantages in industry competition. Especially in the data collection and processing links, accurate and timely acquisition and analysis of data has become an important means to promote the intelligence of the textile industry.

[0003] In the existing technical solutions, data collection is the core link of spinning informatization, and its importance is becoming increasingly prominent. Through technical means such as sensors, spinning companies can obtain various data generated during the operation of equipment in real time and upload them to the cloud for processing and storage.

[0004] However, with the rapid increase in the types and quantity of spinning equipment and the long physical distance between the equipment information communication nodes and the cloud, when a large amount of data is uploaded to the cloud in a short period of time, it is easy to cause network congestion and large delays. The large delay will cause equipment failures to not be discovered in time, which will lead to more serious problems such as production downtime, quality decline, etc., increasing operating and maintenance costs. Summary of the invention

[0005] In order to solve the technical problem that when a large amount of data is uploaded to the cloud in a short period of time, it is easy to cause network congestion and large delay. The large delay will cause equipment failure to be discovered in time, which will further cause more serious problems such as production downtime, quality degradation, etc., and increase operation and maintenance costs, the present invention provides a spinning data collection delay optimization method and system.

[0006] The technical solution provided by the embodiment of the present invention is as follows:

[0007] First aspect:

[0008] A spinning data collection delay optimization method provided by an embodiment of the present invention includes:

[0009] S1: Collect multiple groups of spinning data, and take each group of spinning data as a task unit;

[0010] S2: Determine the edge latency of each task unit when it is processed at the edge and the cloud latency when it is processed in the cloud;

[0011] S3: Determine the total delay of each task unit during processing according to the edge delay and the cloud delay;

[0012] S4: Constructing a spinning data collection delay optimization model;

[0013] S5: setting the constraint conditions of the spinning data collection delay optimization model;

[0014] S6: Under the constraints of the constraints, with the goal of minimizing the total delay, the spinning data collection delay optimization model is optimized by the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan;

[0015] S7: Collect spinning data according to the optimal task allocation plan and the optimal computing resource allocation plan.

[0016] Second aspect:

[0017] A spinning data collection delay optimization system provided by an embodiment of the present invention includes: a memory and one or more processors;

[0018] One or more application programs are stored in the memory, and the one or more application programs are suitable for being executed by the one or more processors to implement the above-mentioned spinning data collection delay optimization method.

[0019] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0020] In the present invention, by constructing a spinning data collection delay optimization model, setting the constraints of the spinning data collection delay optimization model, and under the constraints of the constraints, with the goal of minimizing the total delay, the spinning data collection delay optimization model is optimized through the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan. When a large amount of data is uploaded to the cloud in a short time, it is not easy to cause network congestion and large delays, and equipment failures can be discovered in time, avoiding more serious problems such as production downtime and quality degradation, thereby reducing operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1A schematic flow chart of a spinning data collection delay optimization method provided by an embodiment of the present invention;

[0023] Figure 2 A structural schematic diagram of a spinning data acquisition delay optimization system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0025] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0026] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0027] Reference Manual Attached Figure 1 , showing a flow chart of a spinning data collection delay optimization method provided by an embodiment of the present invention.

[0028] The embodiment of the present invention provides a spinning data collection delay optimization method, which can be implemented by a spinning data collection delay optimization device, and the spinning data collection delay optimization device can be a terminal or a server. The processing flow of the spinning data collection delay optimization method can include the following steps:

[0029] S1: Collect multiple groups of spinning data, and take each group of spinning data as a task unit.

[0030] Optionally, the spinning data includes multiple types:

[0031] Identification attributes: including workshop number, equipment model, equipment number, number of equipment sections and number of spindles per section. These data serve as the only "identity card" for the equipment when it is connected to the network, ensuring the correct communication between the various systems in the factory. If a delay occurs, the equipment may be misidentified or unable to connect to the system in time, which in turn affects production scheduling and management.

[0032] Time attributes: including the start time of the current shift, the current time, and the total equipment operation time. This type of data provides a time dimension reference for the equipment's operation cycle, helping to accurately calculate equipment utilization and maintenance plans. If delays occur, time records may be disrupted, affecting the statistics and analysis of shift production efficiency.

[0033] Shift attributes: including shift selection, end breaks, shift output, and shift efficiency. This type of data is directly related to the production status of each shift. Once a delay occurs, the data will not reflect the real-time production status of the shift, which may cause a delay in adjusting the production plan and affect the overall efficiency of the production line.

[0034] Status attributes: including equipment status, real-time output, yarn type, yarn speed and alarm number. This information determines the real-time operation of the equipment, especially when a fault occurs. The delay of alarm data will affect the timely maintenance of the equipment, resulting in extended downtime, which in turn affects the continuity of production.

[0035] Spindle attributes: including single spindle speed, single spindle status, single spindle output and single spindle efficiency. The amount of data is extremely large. For example, each two-for-one twister has X sections, and each section has Y spindles. Then each device has a total of XY groups of data, which account for the largest proportion of data. If there is a delay, it will lead to deviations in the calculation of spindle output and efficiency, resulting in statistical errors in total output and efficiency, which will seriously affect the accuracy of information management.

[0036] Each spinning data usually occupies 1 or 2 bytes. For example, when the output of a single spindle is 1600, the transmitted hexadecimal data is '0640', and 4 characters occupy 2 bytes. According to statistical analysis, the amount of data transmitted by a device each time is 1.5KB to 2KB.

[0037] Specifically, the overall process of spinning is a series of processes that convert natural fibers (such as cotton, wool, etc.) into continuous filaments or yarns that can be used for weaving. These processes need to be processed by a variety of spinning equipment: First, the cotton is processed by the carding machine and combed into continuous raw cotton strips. Next, multiple cotton strips are combined in the drawing frame and go through the drafting and twisting processes at the same time. Subsequently, the cooked cotton strips produced by the drawing frame are sent to the roving frame for further stretching and twisting. Finally, the final stretching and twisting are carried out by the spinning frame, making the yarn more delicate and tough. If a yarn with a higher twist is required, a double twister is used to twist the preliminarily processed single yarn or ply again to increase its strength and stability. Each spinning device continuously generates spinning data during operation.

[0038] Furthermore, the spinning equipment terminal collects the spinning data generated by the spinning equipment during operation, and packages the collected set of spinning data into an independent data packet, which is defined as a task unit A. i , i=1,2,3…,n:

[0039] A={A 1 ,A 2 ,A 3 ,…,A n}

[0040] Among them, A represents the task unit, A 1 Indicates the first task unit, A 2 Indicates the second task unit, A 3 Indicates the third task unit, A n Indicates n Task unit.

[0041] It should be noted that the spinning equipment terminal is located inside the spinning workshop and is responsible for directly obtaining spinning production data. It includes spinning equipment and data collectors. Spinning equipment is a type of machine that spins cotton fibers into yarn, such as carding machines, spinning machines, and double twisting machines. Each device is equipped with a data collector (also called a collection box) to collect the production and operation data of the equipment and send this data to the edge when specific conditions are met.

[0042] In the present invention, by directly acquiring real-time data from the equipment, it is possible to quickly respond to problems that arise in the production process and adjust the production strategy in a timely manner, thereby reducing production downtime. Acquiring key data such as equipment status, output and efficiency can help production managers monitor production conditions in real time, make quick decisions, and optimize production scheduling. Through accurate data analysis, resource allocation can be optimized, unnecessary waste can be reduced, and production costs can be reduced.

[0043] S2: Determine the edge latency of each task unit when it is processed at the edge and the cloud latency when it is processed in the cloud.

[0044] It should be noted that the edge end is located inside the spinning workshop and consists of multiple edge servers, which are usually deployed close to the spinning equipment terminal. This layout reduces the data transmission distance and time, so that the data from the spinning equipment terminal can be processed quickly and responded to in a timely manner. In addition, the newly added edge end can also reduce the data processing burden of the cloud server. In the present invention, the edge end is responsible for receiving data input from the spinning equipment terminal. Among them, the data assigned to the edge end will be processed on the spot, and the remaining data will be forwarded to the cloud for processing.

[0045] It should be noted that the cloud consists of a cloud server with powerful computing power and fast processing speed. Compared with the edge, the computing power of the cloud is more powerful and stable, but the main disadvantage is that it is far away from the device terminal, resulting in high transmission delay. The cloud is responsible for processing the data received from the edge and for the data processing tasks assigned to the cloud.

[0046] In the present invention, the edge end is deployed at a position close to the spinning equipment terminal, which can significantly reduce the physical distance of data transmission, thereby reducing the transmission delay. This allows the data generated by the equipment to be processed and responded to more quickly, improving overall production efficiency. The edge end processes part of the data, reducing the burden on the cloud server, thereby avoiding the performance bottleneck that may occur in the cloud during peak processing. This load sharing mechanism can improve the processing capacity of the entire system. Sensitive data or real-time data is processed at the edge end, reducing the need to send all data to the cloud and reducing the risk of data leakage during transmission.

[0047] In a possible implementation manner, the edge latency is specifically:

[0048] T i edge =T i edge-calculate +T i ter-edge-tran

[0049] Among them, T i edge represents the edge latency when the i-th task unit is processed at the edge, T i edge-calculate represents the computational delay of the i-th task unit when computing at the edge, T i ter-edge-tran It represents the transmission delay of the i-th task unit from the spinning equipment terminal to the edge end.

[0050] In the present invention, data from the spinning equipment can be processed faster through fast calculation and transmission, ensuring real-time monitoring and fast response to the equipment status during the production process, reducing the downtime of the production line. By processing most of the data at the edge, the amount of data that needs to be transmitted to the cloud is reduced, thereby reducing bandwidth usage and related communication costs.

[0051] Among them, the calculation delay of the task unit when calculating at the edge is specifically:

[0052]

[0053] Among them, α i represents the proportion of the i-th task unit allocated to the edge end, d i represents the amount of spinning data in the i-th task unit, C represents the amount of calculation required to calculate each byte of spinning data, and f i edge Represents the computing resources allocated to the i-th task unit at the edge.

[0054] The transmission delay of the task unit from the spinning equipment terminal to the edge end is as follows:

[0055]

[0056] Where n represents the total number of task units, v ter-edge It indicates the transmission rate of the task unit from the spinning equipment terminal to the edge end.

[0057] In the present invention, by formulating the edge latency in detail, the specific delay of each task unit during processing can be accurately quantified. This helps production managers better understand and manage the response time of the entire production process, thereby improving overall efficiency. By processing data on-site at the edge, real-time data response can be achieved, ensuring that important information such as equipment status and production efficiency can be quickly captured and processed during the production process, improving the timeliness of decision-making. By processing data at the edge and reducing the amount of data transmitted to the cloud, bandwidth usage can be significantly reduced, thereby saving network costs.

[0058] In a possible implementation, the cloud latency is specifically:

[0059] T i cloud =T i cloud-calculate +T i ter-edge-tran +T i edge-cloud-tran

[0060] Among them, T i cloud represents the cloud latency when the i-th task unit is processed in the cloud, T i cloud-calculate represents the computing delay of the i-th task unit when computing in the cloud, T i ter-edge-tran represents the transmission delay of the i-th task unit from the spinning equipment terminal to the edge end, T i edge-cloud-tran It represents the transmission delay of the i-th task unit from the edge to the cloud.

[0061] In the present invention, by clarifying the components of each delay, managers can better analyze the efficiency of cloud processing, identify potential sources of delay, and formulate corresponding optimization strategies. By controlling and optimizing the transmission from the device terminal to the edge and the cloud, the network burden is reduced and the efficiency of network resource utilization is improved. By finely managing the computing and transmission of the cloud, the utilization efficiency of resources can be maximized and the operating cost can be reduced.

[0062] Among them, the computing delay of the task unit when computing in the cloud is specifically:

[0063]

[0064] Among them, β i represents the proportion of the i-th task unit allocated to the cloud, d i represents the amount of spinning data in the i-th task unit, C represents the amount of calculation required to calculate each byte of spinning data, and f i cloud Represents the computing resources allocated to the i-th task unit in the cloud.

[0065] The transmission delay of the task unit from the edge to the cloud is:

[0066]

[0067] Where n represents the total number of task units, v edge-cloud Indicates the transmission rate of task units from the edge to the cloud.

[0068] In the present invention, by formulating the cloud latency, the latency of each task unit in the cloud processing can be accurately evaluated, helping managers to identify and solve potential performance bottlenecks in a timely manner and improve the efficiency of data processing. By accurately calculating the transmission latency from the edge to the cloud, the data transmission path and method can be optimized, unnecessary transmission delays can be reduced, and the overall efficiency of data transmission can be improved. The powerful computing power of the cloud can quickly take over the data processing tasks when problems occur at the edge, ensuring the stability and reliability of the system.

[0069] S3: Determine the total latency of each task unit during processing based on the edge latency and cloud latency.

[0070] In the present invention, by calculating the total delay, the performance of the system can be comprehensively evaluated to understand the overall efficiency when processing each task unit. By analyzing the total delay, it is possible to identify which task units have higher delays, so as to perform reasonable resource allocation and scheduling according to priority, ensuring that key tasks can be completed in a timely manner.

[0071] In a possible implementation manner, the total delay is specifically:

[0072] T i =max(T i edge ,T i clou d)

[0073] Among them, T i represents the total delay of the i-th task unit during processing, max represents the maximum value, T i edge represents the edge latency when the i-th task unit is processed at the edge, T i cloudIt represents the cloud latency when the i-th task unit is processed in the cloud.

[0074] It should be noted that since the task unit is processed simultaneously at the edge and in the cloud, the total latency is the larger value of the edge latency when the task unit is processed at the edge and the cloud latency when it is processed in the cloud.

[0075] In the present invention, since the task units are processed simultaneously at the edge and the cloud, the total latency uses a larger value to more accurately reflect the actual latency required to complete the task. This design ensures that the system can evaluate the processing time in a more realistic way. By comparing the edge latency and the cloud latency, it is possible to identify which part is processed more slowly, thereby providing a basis for optimizing resource allocation. Managers can make improvements and investments in the parts with weaker performance to improve overall efficiency.

[0076] S4: Construct a spinning data collection delay optimization model.

[0077] In the present invention, by establishing an optimization model, each link in the data collection process can be analyzed from a global perspective, possible bottlenecks can be identified, and then systematic optimization can be performed to improve the overall efficiency of data processing. Through the optimization model, computing resources can be dynamically allocated according to the needs of task units to ensure that key tasks obtain sufficient resources while avoiding idleness or waste of resources.

[0078] S5: Setting constraints of the spinning data collection delay optimization model.

[0079] In the present invention, by setting constraints, it is possible to ensure that the solution provided by the optimization model is feasible in actual operation, avoiding unfeasible resource configuration or scheduling solutions. Constraints can ensure the reasonable allocation of computing resources on the edge and cloud, avoid waste or shortage caused by improper resource allocation, and thus improve the efficiency of resource use.

[0080] Optionally, the constraint conditions include: task unit allocation constraints and computing resource allocation constraints.

[0081] The task unit allocation constraints are as follows:

[0082] α i +β i =1

[0083] 0≤α i ≤1,0≤β i ≤1

[0084] Among them, α i represents the proportion of the i-th task unit allocated to the edge, β i It represents the proportion of the i-th task unit allocated to the cloud.

[0085] The specific computing resource allocation constraints are:

[0086] f i edge ≤0.8F edge ,f i cloud ≤0.5F cloud

[0087]

[0088] Among them, f i edge represents the computing resources allocated to the i-th task unit at the edge, F edge represents the total computing resources at the edge, f i cloud represents the computing resources allocated to the i-th task unit in the cloud, F cloud Represents the total computing resources in the cloud.

[0089] Specifically, in the process of computing resource allocation, in order to avoid server problems caused by excessive computing resource usage, the computing resource allocation is constrained. In the process of computing resource allocation, the resources allocated to all tasks on the edge shall not exceed 80% of the total computing resources on the edge, and all allocated resources on the cloud shall not exceed 50% of the total computing resources on the cloud.

[0090] In the present invention, by limiting the computing resource usage of the edge and cloud, it is possible to effectively avoid performance degradation, crashes or failures of the server due to overload, thereby maintaining the stability of the system. Through constraints, computing resources can be reasonably allocated between the edge and the cloud to ensure that key tasks can obtain the required computing power while avoiding idle and wasted resources. Reasonable allocation of resources can reduce data processing delays, improve the efficiency of data collection and processing, and thus improve the overall efficiency of production.

[0091] S6: Under the constraints, with the goal of minimizing the total delay, the spinning data collection delay optimization model is optimized through the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan.

[0092] It should be noted that the PSO-SA algorithm is a hybrid optimization algorithm that combines the particle swarm optimization algorithm (PSO) and the simulated annealing algorithm (SA).

[0093] Among them, the particle swarm optimization algorithm (PSO) is an optimization algorithm based on swarm intelligence, proposed by Kennedy (J.Kennedy) and Eberhart (RCEberhart) in 1995. PSO simulates the foraging behavior of bird flocks and searches for the optimal solution by moving a group of individuals called "particles" in the search space. Each particle represents a potential solution and has a position and speed. During the operation of the algorithm, the particle adjusts its speed and position according to its own experience and the experience of the best particle in the group, so as to move towards the direction of the optimal solution. Specifically, the basic steps of the PSO algorithm include initializing the position and speed of the particle, evaluating the fitness value of each particle (that is, the value of the objective function), and then updating the speed and position of the particle according to the following principles: Each particle adjusts its speed according to its historical optimal position and the global optimal position, and is affected by the inertia weight, individual learning factor and group learning factor. Through iterative updates, the particle swarm gradually converges to the optimal solution. The PSO algorithm has the advantages of simple implementation, fewer parameters and a wide range of application, so it is widely used in function optimization, combinatorial optimization, machine learning, data mining and other fields.

[0094] Among them, the simulated annealing algorithm (SA) is a probabilistic optimization algorithm inspired by the metal annealing process. It is used to find the global optimal solution, especially in large-scale and complex search spaces. The algorithm simulates the process of a substance gradually cooling down after heating, and gradually reduces the energy of the system by controlling the "temperature" to find the lowest energy state, that is, the optimal solution. In the simulated annealing algorithm, the system temperature is high at the beginning, so that the algorithm can accept poor solutions to avoid falling into the local optimum. As the temperature gradually decreases, the probability of the algorithm accepting new solutions also decreases accordingly. This process includes the following steps: first, an initial solution is randomly generated, and then a new solution is generated based on the current solution in each iteration, and its energy change is calculated. If the energy of the new solution is lower than the current solution, the algorithm accepts the new solution. If the energy of the new solution is higher, the new solution is accepted with a certain probability, which is related to the current temperature and energy difference. As the iteration proceeds, the temperature gradually decreases, and the probability of accepting poor solutions decreases, thereby guiding the search to the global optimal solution. The simulated annealing algorithm is widely used in combinatorial optimization, function optimization, constrained optimization and other fields due to its global search capability and simple and easy implementation.

[0095] In the present invention, the PSO-SA algorithm combines the advantages of particle swarm optimization (PSO) and simulated annealing (SA), which can effectively avoid falling into the local optimal solution. PSO searches for multiple potential optimal solutions through swarm intelligence, and SA provides a mechanism to accept poor solutions during the search process to further explore the search space and find the global optimal solution. In the presence of multiple constraints, the PSO-SA algorithm can flexibly respond to complex constraints, ensure that all task allocation and resource allocation constraints are met during the optimization process, and improve the applicability of the model. The PSO-SA algorithm is suitable for real-time optimization applications, and can continuously adjust and optimize task allocation and resource allocation during the production process to improve overall production efficiency.

[0096] In a possible implementation, S6 specifically includes sub-steps S601 and S602:

[0097] S601: Under the constraints of the constraints, with the goal of minimizing the total delay, construct an objective function:

[0098] minT=min{max(T 1 ,T 2 ,T 3 ,……,T n )}

[0099] Where min means taking the minimum value, T means the total delay of the task unit, max means taking the maximum value, T 1 represents the total delay of the first task unit, T 2 represents the total delay of the second task unit, T 3 represents the total delay of the third task unit, T n It represents the total delay of the nth task unit, where n represents the total number of task units.

[0100] Specifically, the total latency is minimized by minimizing the edge latency and cloud latency respectively:

[0101]

[0102]

[0103] Among them, T edge represents the edge latency when the task unit is processed at the edge, T cloud Indicates the cloud latency when the task unit is processed in the cloud.

[0104] In the present invention, by constructing the objective function, the direction of optimization is clearly defined, ensuring that all decisions are made around reducing the total latency of task processing, thereby improving the effectiveness and pertinence of the model. Minimizing the latency at the edge and cloud respectively can achieve more comprehensive system optimization. By identifying and improving the performance of both, the overall processing efficiency is improved to avoid bottlenecks in a certain link. By optimizing the total latency, enterprises can use resources more effectively, reduce additional costs caused by delays, and thus reduce overall operating costs.

[0105] S602: According to the objective function, the spinning data collection delay optimization model is optimized by the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan.

[0106] In the present invention, the PSO (particle swarm optimization) algorithm has the characteristics of fast convergence and can find a near-optimal solution in a short time. Combined with the global search capability of the SA (simulated annealing) algorithm, it can more effectively find the global optimal solution. Through the optimization algorithm, the scientific allocation of edge and cloud computing resources can be achieved, ensuring that key tasks obtain the necessary resource support and improving resource utilization efficiency.

[0107] In a possible implementation, S602 specifically includes sub-steps S6021 to S6028:

[0108] S6021: Input information parameters of the task unit.

[0109] Specifically, the information parameters of the input task unit include the total number of task units n, the amount of spinning data in the i-th task unit d i , the amount of calculation required to calculate each byte of spinning data C, the transmission rate v of the task unit from the spinning equipment terminal to the edge end ter-edge , the transmission rate v of the task unit from the edge to the cloud edge-cloud , the total computing resources at the edge F edge and the total computing resources F in the cloud cloud .

[0110] Furthermore, initialize the computing resources f allocated to the i-th task unit by the edge end i edge and the computing resources f allocated by the cloud to the i-th task unit i cloud .

[0111] S6022: Initialize the parameters of the PSO-SA algorithm.

[0112] Specifically, the parameters of the PSO-SA algorithm are initialized: including the number of particle swarms N, the number of iterations M, the dimension of each particle (i.e., the total number of task units n), the inertia weight w, and the individual learning factor c. 1 , group learning factor c 2 , current temperature T, annealing rate V.

[0113] S6023: Randomly generate an initial particle swarm, where the initial particle swarm includes multiple particles, each particle representing a feasible task allocation scheme or computing resource allocation scheme.

[0114] S6024: Using the objective function as the fitness function, calculating the fitness value of each particle, and taking the position of the particle with the maximum fitness value as the global optimal position.

[0115] S6025: Update the speed and position of the current particle:

[0116]

[0117] in, represents the velocity of the kth particle at the m+1th iteration, w represents the inertia weight, represents the velocity of the kth particle at the mth iteration, c 1 represents the individual learning factor, r 1 Represents the first random number in the interval [0,1], pbest k represents the historical optimal position of the kth particle, represents the position of the kth particle at the mth iteration, c 2 represents the group learning factor, r 2 represents the second random number in the interval [0,1], gbest represents the global optimal position, represents the position of the kth particle at the m+1th iteration.

[0118] S6026: Calculate the fitness value of the updated particle, and compare the fitness value of the updated particle with the fitness value of the current particle to determine whether the fitness value of the updated particle is greater than the fitness value of the current particle. If so, the position of the updated particle is used as the position of the current particle. Otherwise, the acceptance probability of the updated particle is calculated through the annealing probability decision rule in the simulated annealing algorithm, and when the acceptance probability is greater than a random number in the interval [0,1], the position of the updated particle is used as the position of the current particle:

[0119] ΔE=pbestFit-newFit

[0120]

[0121] Among them, ΔE represents the error between the fitness value of the particle at the historical best position and the fitness value of the updated particle, bestFit represents the fitness value of the particle at the historical best position, newFit represents the fitness value of the updated particle, P represents the acceptance probability, and T' represents the current temperature.

[0122] S6027: When the fitness value of the current particle is greater than the fitness value of the particle at the global optimal position, the position of the current particle is used as the global optimal position.

[0123] S6028: Repeat sub-steps S6025 to S6027 until the maximum number of iterations is reached.

[0124] In the present invention, the particle update mechanism of the PSO algorithm is combined with the global search capability of simulated annealing, which helps to avoid falling into the local optimal solution and improves the possibility of finding the global optimal solution. The dynamic mechanism of updating the position and speed of particles can be quickly adjusted according to the performance of the current solution, which improves the adaptability and flexibility of the algorithm. By comparing and updating the fitness value of each particle, it is ensured that the optimization result has higher robustness and can perform well in many situations.

[0125] S7: Collect spinning data according to the optimal task allocation plan and the optimal computing resource allocation plan.

[0126] In the present invention, optimized task allocation and resource configuration are implemented to ensure that the data collection process is more efficient and reduce data processing delay. The optimal computing resource allocation scheme can ensure that each task unit obtains the necessary computing support, avoid resource waste, and improve resource utilization efficiency. By implementing the optimal scheme, the accuracy and timeliness of the data collection process are ensured, and the integrity and reliability of the data are improved.

[0127] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0128] In the present invention, by constructing a spinning data collection delay optimization model, setting the constraints of the spinning data collection delay optimization model, and under the constraints of the constraints, with the goal of minimizing the total delay, the spinning data collection delay optimization model is optimized through the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan. When a large amount of data is uploaded to the cloud in a short time, it is not easy to cause network congestion and large delays, and equipment failures can be discovered in time, avoiding more serious problems such as production downtime and quality degradation, thereby reducing operation and maintenance costs.

[0129] Reference Manual Attached Figure 2, showing a structural schematic diagram of a spinning data acquisition delay optimization system provided by the present invention.

[0130] The present invention further provides a spinning data collection delay optimization system 30 , comprising: a memory 303 and one or more processors 301 .

[0131] One or more application programs are stored in the memory 303, and the one or more application programs are suitable for being executed by the one or more processors 301 to implement the spinning data acquisition delay optimization method described in the method embodiment.

[0132] The spinning data collection delay optimization system 30 includes: a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, for example, via a bus 302.

[0133] The structure of the spinning data collection delay optimization system 30 does not constitute a limitation on the embodiment of the present invention.

[0134] Processor 301 may be a CPU, a general purpose processor, a DSP, an ASIC, an FPGA or other programmable logic device, a transistor logic device, a hardware component or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of the present invention. Processor 301 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0135] The bus 302 may include a path to transmit information between the above components. The bus 302 may be a PCI bus or an EISA bus, etc. The bus 302 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0136] The memory 303 can be a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, or an EEPROM, a CD-ROM or other optical disk storage, an optical disk storage (including a compressed optical disk, a laser disk, an optical disk, a digital versatile disk, a Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited to these.

[0137] It should be noted that the spinning data collection delay optimization system 30 can implement the above-mentioned spinning data collection delay optimization method and can achieve the same or similar technical effects. To avoid repetition, the present invention will not go into details.

[0138] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0139] In the present invention, by constructing a spinning data collection delay optimization model, setting the constraints of the spinning data collection delay optimization model, and under the constraints of the constraints, with the goal of minimizing the total delay, the spinning data collection delay optimization model is optimized through the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan. When a large amount of data is uploaded to the cloud in a short time, it is not easy to cause network congestion and large delays, and equipment failures can be discovered in time, avoiding more serious problems such as production downtime and quality degradation, thereby reducing operation and maintenance costs.

[0140] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program can be loaded and executed by a processor according to the spinning data acquisition delay optimization method described in the first aspect.

[0141] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0142] There are a few points to note:

[0143] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention, and other structures may refer to the general design.

[0144] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of the layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0145] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0146] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A spinning data collection delay optimization method, characterized in that: include: S1: Collect multiple groups of spinning data, and take each group of spinning data as a task unit; S2: Determine the edge latency of each task unit when it is processed at the edge and the cloud latency when it is processed in the cloud; S3: Determine the total delay of each task unit during processing according to the edge delay and the cloud delay; S4: Constructing a spinning data collection delay optimization model; S5: setting the constraint conditions of the spinning data collection delay optimization model; S6: Under the constraints of the constraints, with the goal of minimizing the total delay, the spinning data collection delay optimization model is optimized by the PSO-SA algorithm to obtain the optimal task allocation plan and the optimal computing resource allocation plan; S7: Collect spinning data according to the optimal task allocation plan and the optimal computing resource allocation plan.

2. The spinning data collection delay optimization method according to claim 1, characterized in that: The edge delay is specifically: T i edge =T i edge-calculate +T i ter-edge-tran Among them, T i edge represents the edge latency when the i-th task unit is processed at the edge, T i edge-calculate represents the computational delay of the i-th task unit when computing at the edge, T i ter-edge-tran It represents the transmission delay of the i-th task unit from the spinning equipment terminal to the edge end.

3. The spinning data collection delay optimization method according to claim 2, characterized in that: The calculation delay of the task unit when calculating at the edge is specifically: Among them, α i represents the proportion of the i-th task unit allocated to the edge, d i represents the amount of spinning data in the i-th task unit, C represents the amount of calculation required to calculate each byte of spinning data, and f i edge Represents the computing resources allocated to the i-th task unit at the edge; The transmission delay of the task unit from the spinning equipment terminal to the edge end is specifically: Where n represents the total number of task units, v ter-edge It indicates the transmission rate of the task unit from the spinning equipment terminal to the edge end.

4. The spinning data collection delay optimization method according to claim 1, characterized in that: The cloud latency is specifically: T i cloud =T i cloud-calculate +T i ter-edge-tran +T i edge-cloud-tran Among them, T i cloud represents the cloud latency when the i-th task unit is processed in the cloud, T i cloud-calculate represents the computing delay of the i-th task unit when computing in the cloud, T i ter-edge-tran represents the transmission delay of the i-th task unit from the spinning equipment terminal to the edge end, T i edge-cloud-tran It represents the transmission delay of the i-th task unit from the edge to the cloud.

5. The spinning data collection delay optimization method according to claim 4, characterized in that: The computing delay of the task unit when computing in the cloud is specifically: Among them, β i represents the proportion of the i-th task unit allocated to the cloud, d i represents the amount of spinning data in the i-th task unit, C represents the amount of calculation required to calculate each byte of spinning data, and f i cloud represents the computing resources allocated to the i-th task unit by the cloud; The transmission delay of the task unit from the edge to the cloud is specifically: Where n represents the total number of task units, v edge-cloud Indicates the transmission rate of task units from the edge to the cloud.

6. The spinning data collection delay optimization method according to claim 1, characterized in that: The total delay is specifically: T i =max(T i edge ,T i cloud ) Among them, T i represents the total delay of the i-th task unit during processing, max represents the maximum value, T i edge represents the edge latency when the i-th task unit is processed at the edge, T i cloud It represents the cloud latency when the i-th task unit is processed in the cloud.

7. The spinning data collection delay optimization method according to claim 1, characterized in that: The constraints include: task unit allocation constraints and computing resource allocation constraints; The task unit allocation constraints are specifically: α i +β i ＝1 0≤α i ≤1，0≤β i ≤1 Among them, α i represents the proportion of the i-th task unit allocated to the edge, β i represents the proportion of the i-th task unit allocated to the cloud; The computing resource allocation constraints are specifically: f i edge ≤0.8F edge ,f i cloud ≤0.5F cloud Among them, f i edge represents the computing resources allocated to the i-th task unit at the edge, F edge represents the total computing resources at the edge, f i cloud represents the computing resources allocated to the i-th task unit in the cloud, F cloud Represents the total computing resources in the cloud.

8. The spinning data collection delay optimization method according to claim 1, characterized in that: The S6 specifically includes: S601: Under the constraints of the constraints, with the goal of minimizing the total delay, construct an objective function: minT=min{max(T1,T2,T3,......,T n )) Where min represents the minimum value, T represents the total delay of the task unit, max represents the maximum value, T1 represents the total delay of the first task unit, T2 represents the total delay of the second task unit, T3 represents the total delay of the third task unit, and T n represents the total delay of the nth task unit, and n represents the total number of task units; S602: According to the objective function, the spinning data collection delay optimization model is optimized by PSO-SA algorithm to obtain an optimal task allocation plan and an optimal computing resource allocation plan.

9. The spinning data collection delay optimization method according to claim 8, characterized in that: The S602 specifically includes: S6021: Input information parameters of the task unit; S6022: Initialize the parameters of the PSO-SA algorithm; S6023: Randomly generate an initial particle swarm, where the initial particle swarm includes a plurality of particles, each particle representing a feasible task allocation scheme or computing resource allocation scheme; S6024: using the objective function as a fitness function, calculating the fitness value of each particle, and taking the position of the particle with the maximum fitness value as the global optimal position; S6025: Update the speed and position of the current particle: in, represents the velocity of the kth particle at the m+1th iteration, w represents the inertia weight, represents the speed of the kth particle at the mth iteration, c1 represents the individual learning factor, r1 represents the first random number in the interval [0, 1], and pbest k represents the historical optimal position of the kth particle, represents the position of the kth particle at the mth iteration, c2 represents the group learning factor, r2 represents the second random number in the interval [0, 1], gbest represents the global optimal position, represents the position of the kth particle at the m+1th iteration; S6026: Calculate the fitness value of the updated particle, and compare the fitness value of the updated particle with the fitness value of the current particle to determine whether the fitness value of the updated particle is greater than the fitness value of the current particle; if so, use the position of the updated particle as the position of the current particle; otherwise, calculate the acceptance probability of the updated particle through the annealing probability decision rule in the simulated annealing algorithm, and when the acceptance probability is greater than a random number in the interval [0, 1], use the position of the updated particle as the position of the current particle: ΔE=pbestFit--newFit Among them, ΔE represents the error between the fitness value of the particle at the historical best position and the fitness value of the updated particle, bestFit represents the fitness value of the particle at the historical best position, newFit represents the fitness value of the updated particle, P represents the acceptance probability, and T′ represents the current temperature; S6027: When the fitness value of the current particle is greater than the fitness value of the particle at the global optimal position, the position of the current particle is used as the global optimal position; S6028: Repeat sub-steps S6025 to S6027 until the maximum number of iterations is reached.

10. A spinning data collection delay optimization system, characterized in that: include: memory and one or more processors; One or more application programs are stored in the memory, and the one or more application programs are suitable for being executed by the one or more processors to implement the spinning data collection delay optimization method according to any one of claims 1 to 9.

Images