Concurrent conflict management system and method for OTA batch upgrade

Abstract

The invention discloses a concurrency conflict management system and method for OTA batch upgrade, and relates to the technical field of OTA upgrade.The method comprises the steps that a preset upgrade period is divided through adaptive upgrade time windows, and an adaptive upgrade time window sequence is determined; on the basis of the adaptive upgrade time window sequence, respectively carrying out upgrade execution probability and network connection state prediction on the plurality of terminal devices, and constructing predicted upgrade success probability distribution; performing optimization search by taking the minimum overall upgrading duration and the minimum overall upgrading failure probability as targets, and outputting an optimal upgrading batch layering scheme; and controlling the OTA cloud platform to sequentially issue an upgrading instruction to the terminal equipment in each batch according to the optimal upgrading batch layering scheme, and executing batch upgrading tasks. The problems of upgrading failure, equipment abnormity and even system crash caused by factors such as a large number of terminal equipment, complex and variable network conditions and limited system resources in the existing upgrading technology are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of OTA upgrade, in particular to a concurrent conflict management system and method for OTA batch upgrade. BACKGROUND

[0002] With the rapid development of Internet of Things technology, the number of terminal devices is growing explosively. Terminal devices often need to be remotely upgraded through OTA (Over the Air) technology to fix vulnerabilities, add functions, or improve performance. However, in the batch upgrade process, due to the large number of terminal devices, complex and variable network conditions, limited system resources, and other factors, concurrent conflict problems are easily triggered, leading to upgrade failures, device abnormalities, and even system crashes. The existing technology does not fully consider the concurrent conflict problem, or only optimizes from a single perspective, and cannot effectively solve the complex conflict scenarios in the batch upgrade process. SUMMARY

[0003] The embodiments of the present application provide a concurrent conflict management system and method for OTA batch upgrade, which solves the technical problem that the large number of terminal devices, complex and variable network conditions, and limited system resources easily trigger concurrent conflict problems, leading to upgrade failures, device abnormalities, and even system crashes in the existing upgrade technology.

[0004] The technical solution of the present application to solve the above technical problems is as follows: In a first aspect, the present application provides a concurrent conflict management system for OTA batch upgrade, comprising: A sequence acquisition module is configured to determine an adaptive upgrade time window based on current upgrade task analysis of an OTA cloud platform, divide a preset upgrade period using the adaptive upgrade time window, and determine an adaptive upgrade time window sequence. A probability calculation module is configured to perform upgrade execution probability and network connection state prediction on a plurality of terminal devices based on the adaptive upgrade time window sequence, respectively, obtain a predicted upgrade success probability by fusing the predicted upgrade execution probability and the predicted network connection state, and construct a plurality of predicted upgrade success probability distributions. A scheme optimization module is configured to use the system resource upper limit of the OTA cloud platform as a hard constraint condition, perform upgrade batch hierarchical optimization search based on the plurality of predicted upgrade success probability distributions, with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, and output an optimal upgrade batch hierarchical scheme. A task upgrade module is configured to control the OTA cloud platform to sequentially issue upgrade instructions to terminal devices in each batch according to the optimal upgrade batch hierarchical scheme, and perform batch upgrade tasks.

[0005] In a second aspect, the application provides a concurrent conflict management method for OTA batch upgrade, comprising: determining an adaptive upgrade time window based on current upgrade task analysis of the OTA cloud platform, dividing a preset upgrade period using the adaptive upgrade time window, and determining an adaptive upgrade time window sequence; predicting upgrade execution probability and network connection state of a plurality of terminal devices based on the adaptive upgrade time window sequence, calculating predicted upgrade success probability based on the predicted upgrade execution probability and the predicted network connection state, and constructing a plurality of predicted upgrade success probability distributions; taking the system resource upper limit of the OTA cloud platform as a hard constraint condition, performing upgrade batch hierarchical optimization search based on the plurality of predicted upgrade success probability distributions, taking minimizing overall upgrade time and minimizing overall upgrade failure probability as dual optimization objectives, and outputting an optimal upgrade batch hierarchical scheme; controlling the OTA cloud platform to sequentially issue upgrade instructions to terminal devices in each batch according to the optimal upgrade batch hierarchical scheme, and performing batch upgrade tasks.

[0006] The application provides one or more technical solutions, which have at least the following technical effects or advantages: The application provides a concurrent conflict management system and method for OTA batch upgrade. First, an adaptive upgrade time window is determined based on current upgrade task analysis of the OTA cloud platform, which can more accurately match the best time for device upgrade and avoid upgrade conflicts caused by unsuitable time windows. Second, the adaptive upgrade time window is used to divide a preset upgrade period, and an adaptive upgrade time window sequence is determined, providing an orderly time framework for subsequent upgrade execution. Third, upgrade execution probability and network connection state of a plurality of terminal devices are predicted, and predicted upgrade success probability is calculated by fusion, and a plurality of predicted upgrade success probability distributions are constructed, making the upgrade process more forward-looking and controllable. Fourth, the system resource upper limit of the OTA cloud platform is taken as a hard constraint condition, and predicted upgrade success probability distributions are used to minimize overall upgrade time and minimize overall upgrade failure probability as dual optimization objectives, and upgrade batch hierarchical optimization search is performed to output an optimal upgrade batch hierarchical scheme, effectively solving concurrent conflict problems caused by limited system resources. Finally, the OTA cloud platform is controlled to sequentially issue upgrade instructions to terminal devices in each batch according to the optimal upgrade batch hierarchical scheme, and batch upgrade tasks are performed, improving the success rate and efficiency of upgrade and reducing the risk of upgrade failure, device abnormality, and even system crash.

[0007] By the technical solution, the application embodiment can effectively deal with the OTA batch upgrade concurrency conflict problem caused by the explosive growth of terminal device quantity in the Internet of Things environment. By dynamically adapting the upgrade time window, the difference of network conditions in different periods is fully considered, and the upgrade interruption caused by network fluctuation is avoided. At the same time, by constructing the prediction upgrade success probability distribution, the accurate prediction of device upgrade behavior is realized, and a scientific basis is provided for resource allocation. In terms of resource constraints, the upper limit of system resources is taken as a hard boundary, and hierarchical search is performed in combination with double optimization objectives to ensure the balance between upgrade efficiency and reliability under limited resources. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0009] Figure 1 is a structure schematic diagram of a concurrent conflict management system for OTA batch upgrade provided by the embodiment of the application; Figure 2 is a flow schematic diagram of a concurrent conflict management method for OTA batch upgrade provided by the embodiment of the application.

[0010] In the drawings, the components represented by the numbers are described as follows: Sequence acquisition module 11, probability calculation module 12, scheme optimization module 13, task upgrade module 14. DETAILED DESCRIPTION

[0011] The embodiment of the application provides a concurrent conflict management system and method for OTA batch upgrade, which is used to solve the technical problems that the existing upgrade technology is easy to cause concurrency conflict problem due to the factors such as large number of terminal devices, complex and changeable network conditions and limited system resources, resulting in upgrade failure, device exception and even system crash.

[0012] The technical solutions in the embodiments of the application will be described clearly and completely in combination with the drawings in the embodiments of the application. Obviously, the described embodiments are only some of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0013] In the description of the present application, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can be explicitly or implicitly included one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0014] In the description of the present application, the term "for example" is used to indicate "as an example, illustration or description". Any embodiment described as "for example" in the present application is not necessarily interpreted as more preferred or more advantageous than other embodiments. In order to enable any person skilled in the art to implement and use the present application, the following description is given. In the following description, details are listed for the purpose of explanation. It should be understood that those skilled in the art can recognize that the present application can be implemented without using these specific details. In other examples, well-known structures and processes will not be described in detail to avoid unnecessary details making the description of the present application obscure. Therefore, the present application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope in accordance with the principles and characteristics disclosed.

[0015] Embodiment one, as Figure 1 shown, the embodiment of the present application provides a concurrent conflict management system for OTA batch upgrade, comprising: The sequence acquisition module 11 is configured to determine an adaptive upgrade time window based on the current upgrade task analysis of the OTA cloud platform, divide the preset upgrade period using the adaptive upgrade time window, and determine the adaptive upgrade time window sequence. In the embodiment of the present application, first, the current upgrade task of the OTA cloud platform is analyzed to determine the adaptive upgrade time window, Because under different time windows, network conditions, device load and user usage habits may differ. Through analysis, the time period when the network condition is stable, the device load is low and the user usage frequency is not high is found as the adaptive upgrade time window, so as to minimize the interference and conflict in the upgrade process.

[0016] Secondly, the adaptive upgrade time window is used to divide the preset upgrade period to form an adaptive upgrade time window sequence. The period can be self-defined, such as 15 days. The upgrade time window sequence not only provides a clear time frame for subsequent upgrade execution, but also ensures that the upgrade task can be performed at the best time point, further improving the success rate and efficiency of the upgrade.

[0017] The adaptive upgrade time window is determined based on the current upgrade task analysis of the OTA cloud platform, comprising: collect a sample task attribute information set based on the historical OTA batch upgrade log of the device end, and obtain an upgrade duration mean of a single device under different sample task attribute information as a sample upgrade duration, and obtain a sample upgrade duration set, wherein the task attribute information at least includes an upgrade package volume and an upgrade type; train a machine learning model to convergence using the sample task attribute information set and the sample upgrade duration set, and obtain an upgrade duration predictor; use the upgrade duration predictor to analyze and determine a predicted upgrade duration according to task attribute information of a current upgrade task as an adaptive upgrade time window.

[0018] In the embodiments of the present application, first, based on the device end historical OTA batch upgrade log recording the detailed information of past upgrade tasks, a sample task attribute information set is collected, including attributes such as upgrade package volume, upgrade type, etc. At the same time, the upgrade duration mean of a single device under different sample task attribute information is obtained to form a sample upgrade duration set.

[0019] Among them, the upgrade package volume affects the upgrade duration, and the time required in the transmission and installation process of different volume upgrade packages is different; the upgrade type also affects the upgrade duration, and the complexity and required resources of different types of upgrades in the execution process are different, thereby causing differences in upgrade duration.

[0020] Then, the machine learning model is trained using the sample task attribute information set and the sample upgrade duration set until the model converges, and the upgrade duration predictor is obtained. The upgrade duration predictor can analyze and determine the predicted upgrade duration according to the task attribute information of the current upgrade task, such as the upgrade package volume and the upgrade type, as the basis for adapting the upgrade time window.

[0021] Exemplarily, the upgrade duration predictor is built and trained based on a neural network, and the specific steps are as follows: First, data preparation, collect the sample task attribute information set and the sample upgrade duration set, and collect based on the historical OTA batch upgrade log of the device end; Second, model building, the input layer has a node number equal to the dimension of the input feature, such as the sample task attribute information set and the sample upgrade duration set having a total of 2 features, so the input layer contains 2 nodes; set 1-3 layers of hidden layer, the node number of each layer is adjusted through experiment, such as 64, 32, etc., and the activation function is selected as ReLU; the output layer generally does not use activation function, such as the output time consuming 2 nodes, directly output continuous value.

[0022] Thirdly, the model is trained, and the predicted upgrade duration is taken as the output. The training framework is constructed by using the Adam optimizer and the mean square error loss function, the batch size is set to 32, the total training rounds are set to 50, and the early stopping mechanism (patience = 5, where patience is a hyperparameter for controlling the triggering condition of the early stopping mechanism) is introduced, and when the validation set loss does not decrease for 5 consecutive rounds, the training process is automatically terminated, and the trained upgrade duration predictor is obtained, so as to effectively avoid model overfitting while ensuring that the model reaches a convergent state.

[0023] The upgrade duration predictor is used to determine the adaptive upgrade time window, the network condition and device performance are comprehensively considered to predict the upgrade duration, which can reduce the uncertainty and interference in the upgrade process and improve the success rate and efficiency of the upgrade.

[0024] The probability calculation module 12 is configured to predict the upgrade execution probability and the network connection state of each terminal device based on the adaptive upgrade time window sequence, and to obtain the predicted upgrade success probability by fusing the predicted upgrade execution probability and the predicted network connection state, and to construct a plurality of predicted upgrade success probability distributions. In the embodiment of the application, first, each terminal device is analyzed according to the adaptive upgrade time window sequence. When predicting the upgrade execution probability, the past upgrade records of the device are comprehensively considered.

[0025] For example, based on the number of successful upgrades and the number of failed upgrades in the past, if the device has successfully completed multiple upgrades in the past, the upgrade execution probability will be relatively high. At the same time, the hardware performance of the device, such as processor speed and memory size, is considered. The probability of failure of a device with strong performance in performing an upgrade task is low, and the upgrade execution probability is also high.

[0026] Secondly, for the prediction of the network connection state, the current network environment information and the connection performance of the device under similar network conditions in the past are combined. If the current network signal strength is good and stable, and the device has maintained good connection under the same network strength in the past, the probability of predicting that the network connection state is good will be larger.

[0027] Thirdly, after obtaining the predicted upgrade execution probability and the predicted network connection state, the two are fused and calculated according to the weight ratio. The weights of the upgrade execution probability and the network connection state may be different according to the actual situation. For example, in an environment where the network condition is generally poor, the weight of the network connection state may be appropriately increased. Through fusion calculation, the predicted upgrade success probability of each terminal device is finally obtained.

[0028] Afterwards, the predicted upgrade success probability of all terminal devices is summarized to construct several predicted upgrade success probability distributions. The upgrade success probability distribution can intuitively reflect the possibility of the upgrade success of each terminal device under different conditions, and provide data basis for subsequent scheme optimization.

[0029] Specifically, step 12 in the system comprises: The first time window in the adaptive upgrade time window sequence is selected as the first adaptive upgrade time window, and any device in the several terminal devices is randomly selected as the first terminal device. Based on the first OTA upgrade log and the running state log of the first terminal device, the upgrade execution frequency ratio and the network state quality coefficient mean of the first terminal device in the same historical period are counted as the first predicted upgrade execution probability and the first predicted network connection state, with the first adaptive upgrade time window as the constraint. The first predicted upgrade success probability is calculated based on the fusion of the first predicted upgrade execution probability and the first predicted network connection state, and the first predicted upgrade success probability sequence of the first terminal device in the adaptive upgrade time window sequence is sequentially analyzed as the first predicted upgrade success probability distribution, which is added to the several predicted upgrade success probability distributions.

[0030] In the embodiment of the application, first, the first time window in the adaptive upgrade time window sequence is selected as the first adaptive upgrade time window, and one device in the several terminal devices is randomly selected as the first terminal device. The purpose is to start from a specific time window and a device to start the construction process of the predicted upgrade success probability distribution.

[0031] Secondly, based on the first OTA upgrade log and the running state log of the first terminal device, the upgrade execution frequency ratio of the device in the same historical period is counted, such as the upgrade execution frequency ratio from 9:00 to 9:30 every Wednesday, which reflects the frequency of the device executing the upgrade task in the period, as a reference for the predicted upgrade execution probability.

[0032] At the same time, the network state quality coefficient mean is counted. The network state quality coefficient comprehensively considers the stability, speed and other factors of the network, and the network state quality coefficient mean can reflect the average quality of the network connection of the device in the period, as the first predicted network connection state.

[0033] Then, the predicted upgrade success probability of the first terminal device in other adaptive upgrade time windows is sequentially analyzed in the same way to construct the first predicted upgrade success probability sequence of the first terminal device in the adaptive upgrade time window sequence as the first predicted upgrade success probability distribution.

[0034] Then, based on the first predicted upgrade execution probability and the first predicted network connection state, a fusion calculation is performed according to a preset weight ratio to obtain a first predicted upgrade success probability. The first predicted upgrade success probability comprehensively considers the upgrade execution capability of the device itself and the influence of the network environment on the upgrade.

[0035] Exemplarily, it is assumed that the weight of the upgrade execution probability is 0.6 and the weight of the network connection state is 0.4. If the first predicted upgrade execution probability is 0.8 and the first predicted network connection state is 0.7, the first predicted upgrade success probability obtained through the fusion calculation is 0.6*0.8+0.4*0.7=0.76.

[0036] Then, the first predicted upgrade success probability of the first terminal device in each time window in the adaptive upgrade time window sequence is obtained in sequence according to the same method, a first predicted upgrade success probability sequence is formed, which is used as a first predicted upgrade success probability distribution, and is added to a plurality of predicted upgrade success probability distributions.

[0037] In the above manner, a complete predicted upgrade success probability distribution is gradually constructed, which provides comprehensive and accurate data support for subsequent scheme optimization.

[0038] Further, the first predicted upgrade success probability is obtained based on the fusion calculation of the first predicted upgrade execution probability and the first predicted network connection state, comprising: simulating a task upgrade according to the task attribute information of the current upgrade task and the first predicted network connection state, and outputting a first simulated upgrade time length; half of the adaptive upgrade time window is used as a preset standard upgrade time length; if the first simulated upgrade time length is less than or equal to the preset standard upgrade time length, the first initial upgrade success probability is set to 1, if the first simulated upgrade time length is greater than the preset standard upgrade time length, the ratio of the preset standard upgrade time length to the first simulated upgrade time length is set as the first initial upgrade success probability, and if the first simulated upgrade time length is greater than or equal to the adaptive upgrade time window, the first initial upgrade success probability is set to 0; The first predicted upgrade success probability is obtained by compensating the first initial upgrade success probability according to the first predicted upgrade execution probability.

[0039] In the embodiment of the application, first, the first simulated upgrade time length is obtained by simulating the entire upgrade process according to the task attribute information of the current upgrade task, such as the upgrade package volume, the upgrade type, etc., and the first predicted network connection state, which reflects the approximate time required to complete the upgrade task under the current network conditions.

[0040] Next, half of the adaptation upgrade time window is set as the preset standard upgrade duration, which is used to measure whether the simulated upgrade duration is reasonable.

[0041] Then, based on the comparison between the first simulated upgrade duration and the preset standard upgrade duration, the first initial upgrade success probability is set.

[0042] Furthermore, if the first simulated upgrade duration is less than or equal to the preset standard upgrade duration, it indicates that the upgrade task is likely to be completed smoothly within the adaptation upgrade time window, therefore the first initial upgrade success probability is set to 1. If the first simulated upgrade duration is longer than the preset standard upgrade duration but shorter than the adaptation upgrade time window, the ratio of the preset standard upgrade duration to the first simulated upgrade duration is set as the first initial upgrade success probability to reflect the difficulty of completing the upgrade task. If the first simulated upgrade duration is greater than or equal to the adaptation upgrade time window, it indicates that the upgrade task may not be completed within the adaptation upgrade time window, therefore the first initial upgrade success probability is set to 0.

[0043] Finally, the first initial upgrade success probability is compensated based on the first predicted upgrade execution probability by direct multiplication. For example, if the first initial upgrade success probability is 0.8 and the first predicted upgrade execution probability is 0.9, then the compensated first predicted upgrade success probability is 0.8 × 0.9 = 0.72.

[0044] The first predicted upgrade execution probability reflects the device's execution capability in past upgrades. If the device has a high past upgrade execution probability, it indicates that the device has a strong upgrade execution capability. Therefore, the first initial upgrade success probability can be appropriately increased to obtain the final first predicted upgrade success probability.

[0045] By using the above methods, we can more accurately predict the success rate of each terminal device's upgrade under different adaptation and upgrade time windows, providing support for the optimization of subsequent upgrade solutions.

[0046] The scheme optimization module 13 is used to perform upgrade batch layer optimization search based on the system resource limit of the OTA cloud platform as a hard constraint, and based on the several predicted upgrade success probability distributions, with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, and output the optimal upgrade batch layer scheme. In this embodiment of the application, firstly, the upper limit of the system resources of the OTA cloud platform is defined as a fixed value, representing the maximum amount of resources that the cloud platform can provide during the upgrade process, including but not limited to bandwidth and computing power, as a hard constraint to ensure that the system will not crash or other serious problems due to insufficient resources during the upgrade process.

[0047] Next, based on the previously constructed distributions of predicted upgrade success probabilities, the likelihood of each terminal device successfully upgrading is reflected under different adaptation upgrade time windows. The dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability mean not only completing the upgrade of all devices as quickly as possible, but also ensuring that the number of devices failing during the upgrade process is minimized.

[0048] Next, a layered optimization search for upgrade batches is performed. Specifically, all terminal devices to be upgraded are layered according to certain rules, such as device model, performance, and past upgrade records. Within each layer, different combinations of upgrade batches are attempted based on the predicted upgrade success probability distribution and the system resource limit.

[0049] For example, in some layers, devices with a high probability of successful upgrades are prioritized for upgrades to reduce the overall probability of upgrade failure; while in other layers, upgrade batches are arranged reasonably to make full use of system resources and shorten the overall upgrade time.

[0050] During the search process, the overall upgrade duration and overall upgrade failure probability of each upgrade batch tiered scheme are continuously evaluated. By comparing the objective function values ​​of different schemes, i.e., the weighted sum of the overall upgrade duration and overall upgrade failure probability, the weights can be adjusted according to actual needs to gradually select the better scheme.

[0051] Finally, after searching and comparing, the optimal upgrade batch tiering scheme is output. Under the condition of meeting the system resource limits, this scheme aims to minimize the overall upgrade time and the overall upgrade failure probability, providing an efficient and reliable execution solution for OTA batch upgrades.

[0052] The system resource limits of the OTA cloud platform include network egress bandwidth, server concurrent processing capacity, and database connection pool and I / O load.

[0053] In this embodiment of the application, the system resource limit of the OTA cloud platform ensures the upgrade process. The system resource limit includes network egress bandwidth, server concurrent processing capacity, and database connection pool and I / O load.

[0054] Network egress bandwidth determines the maximum data transmission rate between the cloud platform and terminal devices. Insufficient bandwidth will lead to slow download of upgrade packages, thus prolonging the overall upgrade time. Server concurrency processing capacity reflects the cloud platform's ability to handle multiple upgrade tasks simultaneously. When the concurrency processing capacity reaches its limit, newly added upgrade tasks will not be processed in a timely manner, potentially causing a backlog of upgrade tasks and increasing the probability of overall upgrade failure. Database connection pool and I / O load affect the cloud platform's efficiency in reading and writing upgrade-related data. If database connection pool resources are scarce or I / O load is too high, data query and update operations will slow down, affecting the smoothness of the upgrade process.

[0055] Therefore, when determining the upper limit of system resources, it is necessary to consider factors such as network egress bandwidth, server concurrent processing capacity, database connection pool and I / O load, etc., to ensure that system resources can be reasonably allocated and effectively utilized during the upgrade process, provide reliable constraints for upgrade batch hierarchical optimization search, and thus output the optimal upgrade batch hierarchical scheme that satisfies the dual optimization objectives.

[0056] Specifically, step 13 in the system includes: Using the system resource limit of the OTA cloud platform as a hard constraint, the upper limit of the number of devices to be upgraded in batches is analyzed within the adaptation and upgrade time window, and the upgrade quantity threshold value is output. Based on the adaptation upgrade time window sequence, the upgrade quantity threshold is mapped and compensated according to a preset attenuation ratio sequence, and an adaptation upgrade quantity threshold sequence is output. The preset attenuation ratio is the product of the sequence value of the adaptation upgrade time window and the attenuation step size, and the attenuation step size is 1%. Using the adaptation upgrade quantity threshold sequence as a hierarchical constraint, the upgrade batches of the several terminal devices are randomly set to generate multiple upgrade batch hierarchical schemes. With the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, an upgrade batch stratification optimization search is performed based on the aforementioned several predicted upgrade success probability distributions and multiple upgrade batch stratification schemes, and the optimal upgrade batch stratification scheme is output.

[0057] In this embodiment, firstly, the system resource limit of the OTA cloud platform is used as a hard constraint to analyze the maximum number of devices that can be upgraded in batches within the adaptation and upgrade time window. Given the system resources, the maximum number of devices that can be safely upgraded within each time window is determined to avoid upgrade failures or system crashes due to resource overload, and an upgrade quantity threshold is output.

[0058] Next, based on the adaptation upgrade time window sequence, the upgrade quantity threshold is mapped and compensated according to a preset attenuation ratio sequence to obtain the adaptation upgrade quantity threshold sequence. The preset attenuation ratio is determined by the product of the sequence value of the adaptation upgrade time window and the attenuation step size, with the attenuation step size set to 1%. As system resources may gradually decrease due to other tasks over time, it is necessary to attenuate the upgrade quantity threshold for subsequent time windows, outputting the adaptation upgrade quantity threshold sequence. For example, the entire device population can be divided into multiple levels, and the upgrade scope can be gradually expanded over time, with each batch having a longer reservation time and a larger reservation ratio, forming an inverted triangle shape that gradually widens from top to bottom. This ensures that problems can be detected and contained in the early stages of risk, while ultimately covering all devices.

[0059] Then, using the sequence of upgrade quantity thresholds as hierarchical constraints, the upgrade batches for several terminal devices are randomly set. Multiple hierarchical upgrade batch schemes are generated through random combinations, each scheme satisfying the upgrade quantity limit within different time windows.

[0060] Secondly, with the dual optimization objectives of minimizing the overall upgrade duration and minimizing the overall upgrade failure probability, a layered optimization search for upgrade batches is conducted based on several previously constructed predicted upgrade success probability distributions and multiple upgrade batch stratification schemes. During the search process, the algorithm evaluates the overall upgrade duration and overall upgrade failure probability of each scheme, compares the objective function values ​​of different schemes, and gradually selects the better scheme.

[0061] Finally, after searching and comparing, the optimal upgrade batch tiering scheme was output. This scheme, while meeting the system resource limits, maximizes the goals of minimizing the overall upgrade time and minimizing the overall upgrade failure probability.

[0062] Among them, with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, the upgrade batch stratification optimization search is performed based on the aforementioned several predicted upgrade success probability distributions and multiple upgrade batch stratification schemes, including: Obtain the number of multiple adaptive upgrade time windows corresponding to the multiple upgrade batch layering schemes, and sum the durations to obtain multiple overall upgrade durations; Based on the several predicted upgrade success probability distributions, the upgrade failure probability of each terminal device in the multiple upgrade batch stratification schemes is calculated respectively, and the probabilities are summed to obtain multiple overall upgrade failure probabilities, wherein the upgrade failure probability is the difference between 1 and the upgrade success probability. With the dual optimization objectives of minimizing the overall upgrade duration and minimizing the overall upgrade failure probability, multiple scheme fitnesss are obtained by weighted evaluation based on the multiple overall upgrade durations and multiple overall upgrade failure probabilities, wherein the scheme fitness is negatively correlated with the overall upgrade duration and the overall upgrade failure probability; Based on the fitness of the multiple schemes and the multiple upgrade batch stratification schemes, an upgrade batch stratification optimization search is performed to output the optimal upgrade batch stratification scheme.

[0063] In this embodiment of the application, firstly, the number of multiple adaptation upgrade time windows corresponding to multiple upgrade batch tiered schemes is obtained, and the durations within the time windows are summed to obtain the overall upgrade duration corresponding to each scheme.

[0064] Secondly, based on several constructed predicted upgrade success probability distributions, the upgrade failure probability of each terminal device in multiple upgrade batch tiered schemes is calculated. The upgrade failure probability is obtained by subtracting the upgrade success probability from 1, reflecting the likelihood of each device failing the upgrade under a specific scheme. Then, the upgrade failure probabilities of all terminal devices in each scheme are summed to obtain the overall upgrade failure probability corresponding to each scheme.

[0065] For example, suppose there is an upgrade batch tiered scheme that includes three terminal devices. According to the predicted upgrade success probability distribution, the upgrade success probabilities of the three devices are 0.9, 0.8 and 0.7 respectively, and the upgrade failure probabilities are 0.1, 0.2 and 0.3 respectively. The overall upgrade failure probability of this scheme is 0.1 + 0.2 + 0.3 = 0.6.

[0066] Furthermore, with the dual optimization objectives of minimizing the overall upgrade duration and minimizing the overall upgrade failure probability, a weighted evaluation was performed based on multiple overall upgrade durations and multiple overall upgrade failure probabilities to obtain the fitness of multiple solutions. The fitness of a solution is negatively correlated with the overall upgrade duration and the overall upgrade failure probability, indicating that the shorter the overall upgrade duration and the lower the overall upgrade failure probability, the higher the fitness of the solution.

[0067] Finally, an upgrade batch stratification optimization search is performed based on the fitness of multiple schemes and multiple upgrade batch stratification schemes. During the search process, the algorithm sorts and filters each scheme according to its fitness, gradually approaching the optimal solution. After multiple rounds of iteration and optimization, the optimal upgrade batch stratification scheme is finally output.

[0068] Furthermore, based on the fitness of the multiple schemes and the multiple upgrade batch stratification schemes, an upgrade batch stratification optimization search is performed to output the optimal upgrade batch stratification scheme, including: The upgrade batch stratification scheme is used as the initial solution. Multiple initial solutions are arranged in descending order of fitness to generate an initial solution sequence. The first solution in the initial solution sequence is selected as the optimal solution, and the remaining solutions are selected as the inferior solutions. The optimal solution is used as the direction, and the multiple inferior solutions are adjusted according to a preset optimization step size to obtain multiple updated inferior solutions. The optimal solution and multiple updated inferior solutions are sorted in descending order of fitness. After elimination, P updated inferior solutions are selected and supplemented by randomly selected upgrade batch stratified schemes to generate the first updated solution sequence. Here, P decreases as the number of optimization attempts increases. Based on the first updated solution sequence, continue iterative optimization until the preset number of convergences is reached, and output the optimal solution of the current updated solution sequence as the optimal upgrade batch stratification scheme.

[0069] In this embodiment of the application, firstly, the upgrade batch stratification scheme is used as the initial solution, and the schemes are sorted from largest to smallest fitness to generate an initial solution sequence, and schemes with higher fitness are initially screened out.

[0070] Next, the first solution in the initial solution sequence is selected as the optimal solution, and the remaining solutions are considered inferior solutions. Using the optimal solution as a guide, multiple inferior solutions are adjusted according to a preset optimization step size to obtain multiple updated inferior solutions. The optimization step size is adjusted according to actual needs; for example, the step size can be set as a percentage of the overall upgrade time or the change in the overall upgrade failure probability.

[0071] Then, the optimal solutions and multiple updated inferior solutions are sorted according to their fitness from highest to lowest, and the last P updated inferior solutions are eliminated. The value of P decreases as the number of optimization attempts increases. This dynamic adjustment strategy helps maintain the diversity of solutions in the early stages of optimization, while focusing more on the optimal solution in the later stages. Simultaneously, a randomly selected upgrade batch stratified scheme is used to supplement the solution, generating the first updated solution sequence to maintain solution diversity.

[0072] Subsequently, iterative optimization continues based on the first updated solution sequence. In each iteration, the above-mentioned adjustment, sorting, elimination, and replenishment processes are repeated until a preset number of convergences is reached. The preset number of convergences is determined based on the complexity of the actual upgrade task and the accuracy requirements for the optimal solution. When the preset number of convergences is reached, it means that the algorithm has performed a sufficient number of iterations for optimization. At this point, the optimal solution in the current updated solution sequence is set as the optimal upgrade batch stratification scheme.

[0073] Finally, the optimal solution of the current updated solution sequence is output as the best upgrade batch tiering scheme. Under the condition of satisfying the system resource limit, this scheme achieves the goal of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, providing a reliable execution scheme for OTA batch upgrades.

[0074] The task upgrade module 14 is used to control the OTA cloud platform to send upgrade instructions to the terminal devices in each batch in sequence according to the optimal upgrade batch layering scheme, and to execute the batch upgrade task.

[0075] In this embodiment, after receiving the optimal upgrade batch tiering scheme, the OTA cloud platform is controlled to issue upgrade instructions to the terminal devices in each batch according to the batch order and time window defined in the scheme. Specifically, the upgrade process is initiated batch by batch according to the priority and time schedule set in the scheme, ensuring that each batch of devices can receive timely and stable upgrade services within the limits of system resources.

[0076] Furthermore, during the execution process, the upgrade progress and equipment status are monitored in real time. Once an anomaly is detected, such as equipment response timeout or upgrade failure, the corresponding emergency handling mechanism will be triggered immediately, including retrying the upgrade, recording error logs, and notifying maintenance personnel, thereby ensuring the smooth progress of batch upgrade tasks.

[0077] In summary, compared with existing technologies, this application effectively balances the two mutually restrictive objectives of overall upgrade time and overall upgrade failure probability by optimizing the search through batch upgrade layering. It fully considers the system resource limits of the OTA cloud platform, including network egress bandwidth, server concurrent processing capacity, database connection pool and I / O load, etc., to ensure that system resources can be reasonably allocated and effectively utilized during the upgrade process.

[0078] In summary, the embodiments of this application have at least the following technical effects: This application provides a concurrent conflict management system and method for OTA batch upgrades. First, based on the analysis of current upgrade tasks on the OTA cloud platform, an appropriate upgrade time window is determined, enabling more accurate matching of the optimal time for device upgrades and avoiding upgrade conflicts caused by inappropriate time windows. Second, the preset upgrade cycle is divided using the appropriate upgrade time window to determine an appropriate upgrade time window sequence, providing an orderly time framework for subsequent upgrade execution. Third, by predicting the upgrade execution probability and network connection status of several terminal devices and merging these predictions, a predicted upgrade success probability is obtained, constructing several predicted upgrade success probability distributions. This makes the upgrade process more forward-looking and controllable. Then, using the system resource limit of the OTA cloud platform as a hard constraint, and based on the predicted upgrade success probability distribution, a layered optimization search for upgrade batches is performed with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability. This outputs the optimal upgrade batch layering scheme, effectively solving the concurrency conflict problem caused by limited system resources. Finally, according to the optimal upgrade batch layering scheme, the OTA cloud platform sequentially sends upgrade commands to the terminal devices in each batch, executing batch upgrade tasks, improving the success rate and efficiency of upgrades, and reducing the risk of upgrade failure, device malfunction, or even system crash. Through the above technical solution, this application embodiment can effectively address the OTA batch upgrade concurrency conflict problem caused by the explosive growth in the number of terminal devices in the IoT environment. By dynamically adapting the upgrade time window, the differences in network conditions at different times are fully considered, avoiding upgrade interruptions caused by network fluctuations. Simultaneously, by constructing a predicted upgrade success probability distribution, accurate prediction of device upgrade behavior is achieved, providing a scientific basis for resource allocation. Regarding resource constraints, using the system resource limit as a hard boundary, combined with a layered search based on dual optimization objectives, ensures a balance between upgrade efficiency and reliability under limited resources.

[0079] Example 2, as Figure 2 As shown, based on the same inventive concept as the concurrent conflict management system for OTA batch upgrades provided in Embodiment 1, this application also provides a concurrent conflict management method for OTA batch upgrades, including: Based on the analysis of the current upgrade task on the OTA cloud platform, the appropriate upgrade time window is determined. The preset upgrade cycle is divided using the appropriate upgrade time window to determine the sequence of appropriate upgrade time windows. Based on the adaptation and upgrade time window sequence, the upgrade execution probability and network connection status are predicted for several terminal devices respectively. The predicted upgrade success probability is obtained by fusion calculation of the predicted upgrade execution probability and the predicted network connection status, and several predicted upgrade success probability distributions are constructed. Using the system resource limit of the OTA cloud platform as a hard constraint, and based on the several predicted upgrade success probability distributions, with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, a layered optimization search for upgrade batches is performed, and the optimal upgrade batch layering scheme is output. According to the optimal upgrade batch tiering scheme, the OTA cloud platform is controlled to sequentially issue upgrade instructions to the terminal devices in each batch to execute the batch upgrade task.

[0080] Furthermore, in one embodiment, determining the appropriate upgrade time window based on the current upgrade task analysis of the OTA cloud platform includes: Based on the historical OTA batch upgrade logs of the device, a sample task attribute information set is collected, and the average upgrade time of a single device under different sample task attribute information is obtained as the sample upgrade time, and a sample upgrade time set is obtained. The task attribute information includes at least the upgrade package size and upgrade type. The machine learning model is trained until convergence using the sample task attribute information set and the sample upgrade duration set to obtain the upgrade duration predictor; The upgrade duration predictor is used to analyze and determine the predicted upgrade duration based on the task attribute information of the current upgrade task, which serves as the appropriate upgrade time window.

[0081] Furthermore, based on the aforementioned adaptation and upgrade time window sequence, the probability of upgrade execution and network connection status are predicted for several terminal devices. The predicted success probability of the upgrade is obtained by fusing the predicted upgrade execution probability and the predicted network connection status, and several predicted success probability distributions are constructed, including: The first time window in the adaptation and upgrade time window sequence is selected as the first adaptation and upgrade time window, and any device is randomly selected from the plurality of terminal devices as the first terminal device. Using the first adaptation and upgrade time window as a constraint, based on the first OTA upgrade log and running status log of the first terminal device, the percentage of upgrade execution times and the average network status quality coefficient of the first terminal device in the same historical period are statistically analyzed, which are used as the first predicted upgrade execution probability and the first predicted network connection status. The first predicted upgrade success probability is calculated based on the first predicted upgrade execution probability and the first predicted network connection status. The first predicted upgrade success probability sequence of the first terminal device in the adaptation upgrade time window sequence is then analyzed and added to the plurality of predicted upgrade success probability distributions as the first predicted upgrade success probability distribution.

[0082] Further, in one embodiment, the first predicted upgrade success probability is calculated based on the first predicted upgrade execution probability and the first predicted network connection state, including: Based on the task attribute information of the current upgrade task and the connection status of the first predicted network, the task upgrade simulation is performed, and the first simulated upgrade duration is output. Half of the aforementioned adaptation and upgrade time window is used as the preset standard upgrade duration; If the first simulated upgrade duration is less than or equal to the preset standard upgrade duration, the first initial upgrade success probability is set to 1. If the first simulated upgrade duration is longer than the preset standard upgrade duration, the ratio of the preset standard upgrade duration to the first simulated upgrade duration is set as the first initial upgrade success probability. If the first simulated upgrade duration is greater than or equal to the adaptive upgrade time window, the first initial upgrade success probability is set to 0. The first predicted upgrade success probability is obtained by compensating the first initial upgrade success probability with the first predicted upgrade execution probability.

[0083] The system resource limits of the OTA cloud platform include network egress bandwidth, server concurrent processing capacity, and database connection pool and I / O load.

[0084] Furthermore, in one embodiment, using the system resource limit of the OTA cloud platform as a hard constraint, and based on the several predicted upgrade success probability distributions, an upgrade batch stratification optimization search is performed with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, outputting the optimal upgrade batch stratification scheme, including: Using the system resource limit of the OTA cloud platform as a hard constraint, the upper limit of the number of devices to be upgraded in batches is analyzed within the adaptation and upgrade time window, and the upgrade quantity threshold value is output. Based on the adaptation upgrade time window sequence, the upgrade quantity threshold is mapped and compensated according to a preset attenuation ratio sequence, and an adaptation upgrade quantity threshold sequence is output. The preset attenuation ratio is the product of the sequence value of the adaptation upgrade time window and the attenuation step size, and the attenuation step size is 1%. Using the adaptation upgrade quantity threshold sequence as a hierarchical constraint, the upgrade batches of the several terminal devices are randomly set to generate multiple upgrade batch hierarchical schemes. With the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, an upgrade batch stratification optimization search is performed based on the aforementioned several predicted upgrade success probability distributions and multiple upgrade batch stratification schemes, and the optimal upgrade batch stratification scheme is output.

[0085] Among them, with the dual optimization objectives of minimizing the overall upgrade time and minimizing the overall upgrade failure probability, the upgrade batch stratification optimization search is performed based on the aforementioned several predicted upgrade success probability distributions and multiple upgrade batch stratification schemes, including: Obtain the number of multiple adaptive upgrade time windows corresponding to the multiple upgrade batch layering schemes, and sum the durations to obtain multiple overall upgrade durations; Based on the several predicted upgrade success probability distributions, the upgrade failure probability of each terminal device in the multiple upgrade batch stratification schemes is calculated respectively, and the probabilities are summed to obtain multiple overall upgrade failure probabilities, wherein the upgrade failure probability is the difference between 1 and the upgrade success probability. With the dual optimization objectives of minimizing the overall upgrade duration and minimizing the overall upgrade failure probability, multiple scheme fitnesss are obtained by weighted evaluation based on the multiple overall upgrade durations and multiple overall upgrade failure probabilities, wherein the scheme fitness is negatively correlated with the overall upgrade duration and the overall upgrade failure probability; Based on the fitness of the multiple schemes and the multiple upgrade batch stratification schemes, an upgrade batch stratification optimization search is performed to output the optimal upgrade batch stratification scheme.

[0086] Furthermore, based on the fitness of the multiple schemes and the multiple upgrade batch stratification schemes, an upgrade batch stratification optimization search is performed to output the optimal upgrade batch stratification scheme, including: The upgrade batch stratification scheme is used as the initial solution. Multiple initial solutions are arranged in descending order of fitness to generate an initial solution sequence. The first solution in the initial solution sequence is selected as the optimal solution, and the remaining solutions are selected as the inferior solutions. The optimal solution is used as the direction, and the multiple inferior solutions are adjusted according to a preset optimization step size to obtain multiple updated inferior solutions. The optimal solution and multiple updated inferior solutions are sorted in descending order of fitness. After elimination, P updated inferior solutions are selected and supplemented by randomly selected upgrade batch stratified schemes to generate the first updated solution sequence. Here, P decreases as the number of optimization attempts increases. Based on the first updated solution sequence, continue iterative optimization until the preset number of convergences is reached, and output the optimal solution of the current updated solution sequence as the optimal upgrade batch stratification scheme.

[0087] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0088] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0089] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A concurrent conflict management system for OTA bulk upgrade, characterized in that, The method comprises the following steps: A sequence acquisition module is configured to determine an adaptive upgrade time window based on current upgrade task analysis of an OTA cloud platform, divide a preset upgrade period using the adaptive upgrade time window, and determine an adaptive upgrade time window sequence; A probability calculation module is configured to predict upgrade execution probability and network connection state of a plurality of terminal devices based on the adaptive upgrade time window sequence, calculate predicted upgrade success probability based on the predicted upgrade execution probability and the predicted network connection state, and construct a plurality of predicted upgrade success probability distributions; A scheme optimization module is configured to use the system resource upper limit of the OTA cloud platform as a hard constraint condition, use the plurality of predicted upgrade success probability distributions as a double optimization target of minimizing overall upgrade time and minimizing overall upgrade failure probability, perform upgrade batch hierarchical optimization search, and output an optimal upgrade batch hierarchical scheme; A task upgrade module is configured to control the OTA cloud platform to sequentially issue upgrade instructions to terminal devices in each batch according to the optimal upgrade batch hierarchical scheme, and perform batch upgrade tasks.

2. The concurrent conflict management system for OTA bulk upgrade of claim 1, wherein, The adaptive upgrade time window is determined based on current upgrade task analysis of an OTA cloud platform, comprising: Based on historical OTA batch upgrade logs of the device side, a sample task attribute information set is collected, and the upgrade time length mean of a single device under different sample task attribute information is obtained as a sample upgrade time length, and a sample upgrade time length set is obtained, wherein the task attribute information at least includes upgrade package volume and upgrade type; The machine learning model is trained using the sample task attribute information set and the sample upgrade time length set until convergence, and an upgrade time length predictor is obtained; The upgrade time length predictor is used to analyze and determine the predicted upgrade time length according to the task attribute information of the current upgrade task, as the adaptive upgrade time window.

3. The concurrent conflict management system for OTA bulk upgrade of claim 1, wherein, The upgrade execution probability and network connection state of a plurality of terminal devices are predicted based on the adaptive upgrade time window sequence, the predicted upgrade success probability is calculated based on the predicted upgrade execution probability and the predicted network connection state, and a plurality of predicted upgrade success probability distributions are constructed, comprising: The first adaptive upgrade time window is selected from the adaptive upgrade time window sequence as the first adaptive upgrade time window, and any device in the plurality of terminal devices is randomly selected as the first terminal device; The first adaptive upgrade time window is used as a constraint, the first OTA upgrade log and the running state log of the first terminal device are used to statistically obtain the upgrade execution frequency ratio and the network state quality coefficient mean of the first terminal device in the historical same period as the first predicted upgrade execution probability and the first predicted network connection state; The first predicted upgrade success probability is calculated based on the first predicted upgrade execution probability and the first predicted network connection state, and the first predicted upgrade success probability sequence of the first terminal device in the adaptive upgrade time window sequence is sequentially analyzed and obtained as the first predicted upgrade success probability distribution, which is added to the plurality of predicted upgrade success probability distributions.

4. The concurrent conflict management system for OTA bulk upgrade of claim 3, wherein, obtaining a first predicted upgrade success probability based on the first predicted upgrade execution probability and the first predicted network connection state fusion calculation, comprising: performing a task upgrade simulation according to the task attribute information of the current upgrade task and the first predicted network connection state, and outputting a first simulated upgrade time length; taking one half of the adaptive upgrade time window as a preset standard upgrade time length; if the first simulated upgrade time length is less than or equal to the preset standard upgrade time length, setting a first initial upgrade success probability to 1, if the first simulated upgrade time length is greater than the preset standard upgrade time length, setting a ratio of the preset standard upgrade time length to the first simulated upgrade time length as the first initial upgrade success probability, and if the first simulated upgrade time length is greater than or equal to the adaptive upgrade time window, setting the first initial upgrade success probability to 0; compensating the first initial upgrade success probability according to the first predicted upgrade execution probability to obtain a first predicted upgrade success probability.

5. The concurrent conflict management system for OTA bulk upgrade of claim 1, wherein, The system resource upper limit of the OTA cloud platform includes network export bandwidth, server concurrent processing capability, and database connection pool and I / O load.

6. The concurrent conflict management system for OTA bulk upgrade of claim 5, wherein, Taking the system resource upper limit of the OTA cloud platform as a hard constraint condition, based on the plurality of predicted upgrade success probability distributions, taking minimizing the overall upgrade time length and minimizing the overall upgrade failure probability as dual optimization objectives, performing upgrade batch hierarchical optimization search, and outputting an optimal upgrade batch hierarchical scheme, comprising: Taking the system resource upper limit of the OTA cloud platform as a hard constraint condition, performing device batch upgrade quantity upper limit analysis within the adaptive upgrade time window, and outputting an upgrade quantity threshold value; Based on the adaptive upgrade time window sequence, mapping and compensating the upgrade quantity threshold value according to a preset attenuation ratio sequence, and outputting an adaptive upgrade quantity threshold value sequence, wherein the preset attenuation ratio is a product of the order value of the adaptive upgrade time window and the attenuation step length, and the attenuation step length is 1%. Taking the adaptive upgrade quantity threshold value sequence as a hierarchical constraint, randomly setting the upgrade batch of the plurality of terminal devices to generate a plurality of upgrade batch hierarchical schemes; Taking minimizing the overall upgrade time length and minimizing the overall upgrade failure probability as dual optimization objectives, based on the plurality of predicted upgrade success probability distributions and the plurality of upgrade batch hierarchical schemes, performing upgrade batch hierarchical optimization search, and outputting an optimal upgrade batch hierarchical scheme.

7. The concurrent conflict management system for OTA bulk upgrade of claim 6, wherein, Taking the system resource upper limit of the OTA cloud platform as a hard constraint condition, performing device batch upgrade quantity upper limit analysis within the adaptive upgrade time window, and outputting an upgrade quantity threshold value; Based on the adaptive upgrade time window sequence, mapping and compensating the upgrade quantity threshold value according to a preset attenuation ratio sequence, and outputting an adaptive upgrade quantity threshold value sequence, wherein the preset attenuation ratio is a product of the order value of the adaptive upgrade time window and the attenuation step length, and the attenuation step length is 1%. Taking the adaptive upgrade quantity threshold value sequence as a hierarchical constraint, randomly setting the upgrade batch of the plurality of terminal devices to generate a plurality of upgrade batch hierarchical schemes; Taking minimizing the overall upgrade time length and minimizing the overall upgrade failure probability as dual optimization objectives, based on the plurality of predicted upgrade success probability distributions and the plurality of upgrade batch hierarchical schemes, performing upgrade batch hierarchical optimization search, and outputting an optimal upgrade batch hierarchical scheme. Taking the system resource upper limit of the OTA cloud platform as a hard constraint condition, performing device batch upgrade quantity upper limit analysis within the adaptive upgrade time window, and outputting an upgrade quantity threshold value; Based on the adaptive upgrade time window sequence, mapping and compensating the upgrade quantity threshold value according to a preset attenuation ratio sequence, and outputting an adaptive upgrade quantity threshold value sequence, wherein the preset attenuation ratio is a product of the order value of the adaptive upgrade time window and the attenuation step length, and the attenuation step length is 1%. Taking the adaptive upgrade quantity threshold value sequence as a hierarchical constraint, randomly setting the upgrade batch of the plurality of terminal devices to generate a plurality of upgrade batch hierarchical schemes; Taking minimizing the overall upgrade time length and minimizing the overall upgrade failure probability as dual optimization objectives, based on the plurality of predicted upgrade success probability distributions and the plurality of upgrade batch hierarchical schemes, performing upgrade batch hierarchical optimization search, and outputting an optimal upgrade batch hierarchical scheme. Minimizing the overall upgrade duration and minimizing the overall upgrade failure probability are dual optimization objectives, and a plurality of scheme fitnesses are obtained by weighted evaluation according to the plurality of overall upgrade durations and the plurality of overall upgrade failure probabilities, wherein the scheme fitness is negatively correlated with the overall upgrade duration and the overall upgrade failure probability; Based on the plurality of scheme fitnesses and a plurality of upgrade batch hierarchical schemes, upgrade batch hierarchical optimization search is performed, and an optimal upgrade batch hierarchical scheme is output.

8. The concurrent conflict management system for OTA bulk upgrade of claim 7, wherein, Based on the plurality of scheme fitnesses and a plurality of upgrade batch hierarchical schemes, upgrade batch hierarchical optimization search is performed, and an optimal upgrade batch hierarchical scheme is output, including: The upgrade batch hierarchical scheme is taken as an initial solution, the plurality of initial solutions are arranged in descending order of scheme fitness, and an initial solution sequence is generated; The first solution of the initial solution sequence is set as a superior solution, and the remaining solutions are set as inferior solutions, and the superior solution is taken as a direction, and the plurality of inferior solutions are adjusted according to a preset optimization step, and a plurality of updated inferior solutions are obtained; The superior solution and the plurality of updated inferior solutions are sorted in descending order of scheme fitness, the last P updated inferior solutions are eliminated, and an upgrade batch hierarchical scheme is randomly selected for supplementation, and a first updated solution sequence is generated, wherein P decreases with the increase of the optimization number; Based on the first updated solution sequence, iterative optimization is continued until a preset convergence number is reached, and the superior solution of the current updated solution sequence is output as the optimal upgrade batch hierarchical scheme.

9. A method for concurrent conflict management for OTA bulk upgrade, characterized in that, A concurrent conflict management system for OTA batch upgrade is applied to any one of claims 1-8, comprising: Based on the current upgrade task analysis of the OTA cloud platform, an adaptive upgrade time window is determined, and the adaptive upgrade time window is used to divide the preset upgrade period to determine an adaptive upgrade time window sequence; Based on the adaptive upgrade time window sequence, the upgrade execution probability and the network connection state of a plurality of terminal devices are predicted respectively, the predicted upgrade success probability is obtained by fusion calculation according to the predicted upgrade execution probability and the predicted network connection state, and a plurality of predicted upgrade success probability distributions are constructed; Taking the system resource upper limit of the OTA cloud platform as a hard constraint condition, based on the plurality of predicted upgrade success probability distributions, minimizing the overall upgrade duration and minimizing the overall upgrade failure probability are dual optimization objectives, and upgrade batch hierarchical optimization search is performed, and an optimal upgrade batch hierarchical scheme is output; According to the optimal upgrade batch hierarchical scheme, the OTA cloud platform is controlled to sequentially issue upgrade instructions to the terminal devices in each batch to perform batch upgrade tasks.

Images