A multi-device cooperative control method based on an open-source honkong system development experiment box

Abstract

The present application belongs to the technical field of cooperative control, and particularly relates to a multi-device cooperative control method for a development experiment box based on an open-source Hongmeng system. The method realizes the access and identity authentication, automatic discovery and dynamic connection of the experiment box devices through the distributed device management platform of the open-source Hongmeng system, completes resource information collection, strategy matching and dynamic distribution by using a unified resource scheduling mechanism, performs task allocation in combination with the device state and task demand by using an intelligent task scheduling algorithm, realizes real-time data transmission between devices by means of a low-delay communication protocol, and finally automatically summarizes and analyzes experimental data and generates a report. The method solves the problems of low device identification efficiency, low resource utilization rate and inefficient task allocation in the existing development experiment box, improves the intelligent level and system stability of device cooperative control, and is suitable for the multi-device cooperative control scene of the development experiment box.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of cooperative control, and particularly relates to a multi-device cooperative control method for a development experiment box based on an open-source Hongmeng system. BACKGROUND

[0002] There are many limitations in the multi-device cooperative control of existing development experiment box teaching equipment, which is difficult to meet the modern intelligent teaching and experimental requirements. Most of the traditional systems adopt a static connection and manual configuration mode, the device recognition efficiency is low, the resource utilization rate is not high, there is a lack of unified scheduling mechanism, and it is difficult to realize dynamic allocation and real-time response of experimental tasks. In addition, most systems lack support for device identity security authentication, resource dynamic management and task intelligent scheduling, which can easily cause device conflicts, data asynchronization, control failure and other problems, which seriously restricts the cooperative ability and intelligent level of the teaching experiment environment. With the development of intelligent operating systems, especially the advantages of open-source Hongmeng system in distributed device management and low-latency communication are increasingly prominent, which provides a new technical path for multi-device cooperative control. SUMMARY

[0003] The application proposes a multi-device cooperative control method for a development experiment box based on an open-source Hongmeng system to solve the technical problems in the above background art.

[0004] In order to achieve the above purpose, the technical scheme adopted by the application comprises the following steps:

[0005] S1, device access and identity authentication: through the distributed device management platform in the Hongmeng operating system, multiple devices in the experiment box are accessed and authenticated, and each device is assigned a unique identifier;

[0006] S2, device automatic discovery and dynamic connection: automatically identify the newly added or changed devices in the experiment box, and realize the dynamic connection between the devices;

[0007] S3, device resource sharing: in the open-source Hongmeng system, through a unified resource scheduling mechanism, the sharing and reasonable allocation of the resources of each device in the experiment box are realized, including: resource information collection, scheduling strategy matching and dynamic resource allocation;

[0008] Information collection: through the open-source Hongmeng system interface, the performance parameters and real-time load information of each device in the experiment box are collected;

[0009] Scheduling strategy matching: according to the collected performance parameters and real-time load information, combined with the resource requirements of the experimental tasks, the matching resource allocation strategy is matched;

[0010] Dynamic resource allocation: according to the matched strategy, through the unified scheduling of the open-source Hongmeng system, the idle device resources are allocated to the experimental tasks in need.

[0011] S4, task scheduling and allocation: according to the experimental requirements and equipment state information, an intelligent task scheduling algorithm based on the HONGBAO operating system is adopted to reasonably allocate the experimental tasks to each device, and to ensure that each device can process in real time according to the task priority and cooperative relationship;

[0012] S5, real-time data transmission: using the low-delay communication protocol of the open-source HONGBAO system, the data synchronization and streaming transmission between devices are ensured, the data state between devices is updated in real time, and the data delay and synchronization conflict are reduced;

[0013] S6, experimental data analysis and report generation: after the completion of the experimental task, the system automatically summarizes and analyzes the device data, generates an experimental report based on the real-time task execution result, and displays the analysis result to the user through a visual interface, supporting the user to further optimize the experimental process and make decisions.

[0014] As preferred, the identity authentication in step S1 includes an identification rule definition step, an information collection step, an authentication verification step, and an identifier allocation step for each device;

[0015] Identification rule definition: according to the HONGBAO distributed device management specification, combined with the device type and functional attribute of the experimental box, the unique identifier coding rule is formulated, and the composition structure of the device classification code, function code and serial number is determined;

[0016] Information collection: through the device access interface, the hardware characteristics, device type and manufacturer basic information of each device in the experimental box are collected as the basis for authentication and identification;

[0017] Authentication verification: using the open-source HONGBAO system security authentication framework, the collected device information is compared with the preset device whitelist and security feature library to verify the legality and integrity of the device;

[0018] Identifier allocation: for the authenticated device, a unique identifier is generated and allocated according to the defined identification rule, and the device basic information and authentication status are associated and synchronized into the distributed device management platform identification database.

[0019] As preferred, a dynamic identifier verification and update process is added after the identifier allocation step, including a verification trigger judgment step, an effectiveness determination step, and an identifier update step;

[0020] Verification trigger judgment: set a regular verification period or trigger the identifier validity verification process when the device switches the type of task execution;

[0021] Effectiveness determination: compare the current hardware characteristics and running state parameters of the device with the initial authentication information associated with the identifier to determine whether the device has the risk of hardware tampering and identity impersonation.

[0022] Identifier update: if it is determined that there is a risk, re-perform the information collection and authentication verification steps, generate a new identifier to replace the existing identifier; if it is determined that there is no risk, maintain the existing identifier and associated information.

[0023] As a preferred embodiment, the step S2 device automatic discovery and dynamic connection: automatically identify the newly added or changed devices in the experimental box, and realize the specific implementation operation of dynamic connection between devices:

[0024] Through the distributed software bus of the open source Hongmeng system, multi-dimensional feature extraction is performed on the broadcast signal of the devices in the physical space of the experimental box, and a time-space domain device fingerprint is generated: , wherein are the time domain feature, the spatial feature, the radio frequency feature and the protocol feature, and m represents the index of the device;

[0025] A dynamic topology graph of the experimental box devices is constructed.

[0026] Based on the device fingerprint similarity matrix S, the fingerprint similarity of devices m and n is calculated, and the calculation method is: According to the similarity threshold, an adjacency matrix is dynamically generated, and if the similarity of two devices is greater than the threshold, a connection relationship is established, that is, , wherein is the adjacency matrix, represents the similarity threshold;

[0027] A sliding event window is designed , the time series difference analysis of the device state is performed, and the device change is detected: , wherein is the state vector of the device m at time t, and when the change flag meets the determination condition, it is considered that the existing device state changes, and the determination of the change flag is , wherein is the change sensitivity parameter;

[0028] When a new device joins or the existing device state changes, the topology graph is updated in real time.

[0029] As a preferred embodiment, the scheduling strategy matching in step S3: according to the collected performance parameters and real-time load information, combined with the demand of the experimental task for resources, the specific implementation steps of matching the adaptive resource allocation strategy are:

[0030] The collected device performance parameters and real-time load information are used to construct a device resource state vector , wherein, represents the normalized parameter of the i-th type of resource;

[0031] The demand of the experimental task for resources is synchronously analyzed, and a task demand vector is generated wherein the demand weight of the i-th resource;

[0032] the similarity between the resource vector of the computing device and the demand vector of the task and filter out the device set with similarity greater than the set threshold as the candidate resource pool;

[0033] Based on the candidate resource pool, the final allocation strategy is generated by weighted summation, and the calculation method is: wherein P represents the final allocation strategy, is the resource state vector of the m-th candidate device, is the stability weight of the m-th device performing the task, is the resource load rate of the device m.

[0034] As preferred, between the S3 scheduling strategy matching and the dynamic resource allocation, it is also necessary to check whether the candidate device will appear overload or resource conflict after allocating resources according to the strategy, and if the verification fails, the strategy is adjusted again.

[0035] As preferred, the step S4 task scheduling and allocation: according to the experimental demand and the device state information, the intelligent task scheduling algorithm based on the Hongmeng operating system is adopted to reasonably allocate the experimental tasks to each device, and ensure that each device can process in real time according to the task priority and the cooperative relationship. The specific implementation is:

[0036] Firstly, collect all the experimental task information to be scheduled at present, and construct the task description matrix wherein is the conversion demand intensity of the j-th experimental task on the i-th resource, is the total number of tasks,

[0037] Collect the current running state parameters of each experimental device, and construct the device resource capability matrix wherein is the remaining amount of the i-th resource that the m-th device can provide at present, and M is the total number of devices;

[0038] wherein, is the priority weight of the i-th resource, is the resource matching efficiency of task j allocated to device m, and the match function represents the matching degree function between the task and the device;

[0039] Construct the bipartite graph of the task device, and the edge weight is and introduce the task priority weight ​​​, device load weight , solving the optimization objective function: , wherein , indicates whether the task is assigned to the device.

[0040] As preferred, the objective function needs to meet: each task is assigned to only one device: ; and the total resources of each device cannot be overloaded: .

[0041] Compared with the prior art, the advantages and positive effects of the present application are:

[0042] 1. The device automatically accesses, identity authentication and dynamic identifier management are realized through the distributed device management platform of the Hongmeng, solving the problem of low device recognition efficiency caused by traditional static connection and manual configuration. The dynamic verification mechanism can detect the risk of hardware tampering in real time, ensure the uniqueness and security of device identity, and improve the system attack resistance.

[0043] 2. A unified resource scheduling mechanism is constructed, combining the similarity matching of device resource state vector and task demand vector, realizing resource dynamic allocation and overload conflict checking.

[0044] 3. The data stream transmission between devices is realized by using the low-delay communication protocol of Hongmeng, and the intelligent task scheduling algorithm based on bipartite graph optimization is used, solving the problems of data synchronization conflict and low efficiency of task allocation in traditional systems. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0046] Figure 1 It is a structure flow chart of a multi-device collaborative control method based on an open source Hongmeng system development experiment box. DETAILED DESCRIPTION

[0047] In order to more clearly understand the above-mentioned purposes, features and advantages of the present application, the following will further illustrate the present application by combining with the drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0048] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, however, the present application can also be implemented in other ways different from those described herein, therefore, the present application is not limited to the specific embodiments disclosed in the following description.

[0049] In the embodiment, the existing experimental box equipment cooperative control scheme has significant defects. The mode of static connection and manual configuration leads to low equipment identification efficiency, long-term low resource utilization rate, lack of unified scheduling mechanism, and difficulty in realizing dynamic allocation and real-time response of experimental tasks. Therefore, the present application proposes a development experimental box multi-device cooperative control method based on an open source Hongmeng system, and the specific implementation method is as shown in Figure 1 .

[0050] First, in order to ensure the safety and uniqueness of the equipment access in the experimental box, a distributed equipment management platform in the Hongmeng operating system is used to build a perfect equipment identity authentication system for equipment access and identity authentication: through the distributed equipment management platform in the Hongmeng operating system, multiple devices in the experimental box are accessed and authenticated, and each device is assigned a unique identifier. Specifically, the identity verification assigns a unique identifier to each device, including an identification rule definition step, an information collection step, an authentication verification step, and an identifier allocation step. Specifically, the identification rule defines: according to the Hongmeng distributed equipment management specification, combined with the experimental box equipment type, functional attribute, the unique identifier coding rule is formulated, and the device classification code, function code, and serial number composition structure are clearly defined; information collection: through the device access interface, the hardware characteristics, device type, and manufacturer basic information of each device in the experimental box are collected as the basis for authentication and identification; authentication verification: using the open source Hongmeng system security authentication framework, the collected device information is compared with the preset device whitelist and security feature library to verify the legality and integrity of the device; identifier allocation: for the authenticated device, a unique identifier is generated and allocated according to the defined identification rule, and the device basic information and authentication status are associated and synchronized into the distributed equipment management platform identification database.

[0051] A dynamic identifier verification and update process is added after the identifier allocation step, including a verification trigger judgment step, an effectiveness determination step, and an identifier update step. Specifically, the verification trigger judgment: set a regular verification period or trigger the identifier validity verification process when the device executes task type switching; effectiveness determination: extract the current hardware characteristics and running state parameters of the device, and compare them with the initial authentication information associated with the identifier to determine whether the device has hardware tampering or identity impersonation risk; identifier update: if it is determined that there is a risk, the information collection and authentication verification steps are executed again to generate a new identifier to replace; if it is determined that there is no risk, the existing identifier and associated information are maintained, and the verification result is recorded to the management platform.

[0052] In order to realize the automatic identification and dynamic connection of new or changed devices in the experimental box, and improve the flexibility and adaptability of the system, the distributed soft bus technology of open source Hongmeng system is used to construct the device automatic discovery and dynamic connection mechanism, automatically identify the new or changed devices in the experimental box, and realize the dynamic connection between devices. The specific implementation operation is as follows: through the distributed soft bus of open source Hongmeng system, multi-dimensional feature extraction is carried out on the broadcast signal of the device in the physical space of the experimental box, and the space-time domain device fingerprint is generated: , wherein are time domain feature, space feature, radio frequency feature and protocol feature respectively, and m represents the index of the device; the dynamic topology graph of the experimental box device is constructed; based on the device fingerprint similarity matrix S, the fingerprint similarity of devices m and n is calculated, and the calculation method is: , the adjacency matrix is dynamically generated according to the similarity threshold value, and if the similarity of two devices is greater than the threshold value, the connection relationship is established, that is , wherein is the adjacency matrix, represents the similarity threshold value; a sliding event window is designed to perform time series difference analysis on the device state to detect device changes: , wherein is the state vector of device m at time t, and when the change flag meets the judgment condition, it is considered that the existing device state changes, and the judgment of the change flag is , wherein is the change sensitivity parameter; when a new device joins or the existing device state changes, the topology graph is updated in real time. Ensure that each device can quickly and accurately establish a connection with other devices. In this process, the system can adjust the connection relationship between devices in real time, avoiding the problem of control failure or data asynchronization caused by device changes. Therefore, the implementation of this step makes the experimental system have high automation and flexibility, can quickly respond to environmental changes, effectively improves the efficiency of device management, reduces human operation errors and time consumption, and optimizes the collaborative efficiency of the whole experimental process.

[0053] Next, in order to realize the sharing and reasonable allocation of each device resource in the experimental box and improve the resource utilization rate, a unified resource scheduling mechanism is constructed in the open source Hongmeng system, which specifically includes three links of resource information collection, scheduling strategy matching and dynamic resource allocation. Information collection: through the interface of open source Hongmeng system, the performance parameters and real-time load information of each device in the experimental box are collected; scheduling strategy matching: according to the collected performance parameters and real-time load information, combined with the demand of experimental task for resources, the adaptive resource allocation strategy is matched;

[0054] Dynamic resource allocation: according to the matched strategy, idle device resources are allocated to the experimental tasks in need through unified scheduling of the open source HOMER system. The specific implementation steps of the scheduling strategy matching are as follows: the device performance parameters and real-time load information collected are used to construct a device resource state vector wherein, represents the normalized parameter of the ith resource; the resource demand of the experimental task is analyzed synchronously to generate a task demand vector wherein represents the demand weight of the ith resource; the similarity between the device resource vector and the task demand vector is calculated , and the device set with a similarity greater than a set threshold is selected as a candidate resource pool; based on the candidate resource pool, the final allocation strategy is generated through weighted summation, and the calculation method is as follows: wherein, P represents the final allocation strategy, is the resource state vector of the mth candidate device, represents the stability weight of the mth device for executing the task, wherein represents the resource load rate of the device m. After the final allocation strategy, idle device resources are allocated to the experimental tasks in need through unified scheduling of the open source HOMER system. The device resource sharing step, starting from the resource supply side, dynamically allocates resources according to the task demand, and solves the problem of how to efficiently use resources.

[0055] Between the scheduling strategy matching and the dynamic resource allocation, it is also necessary to check whether the candidate device will appear overload or resource conflict after allocating resources according to the strategy, and if the verification fails, the strategy is adjusted again. First, the current resource state of the candidate device is collected in real time through the open source HOMER system interface, including CPU utilization, memory remaining amount, storage bandwidth and other performance parameters, and a device real-time resource vector is constructed. Second, based on the preliminary allocation strategy, the resource demand of the target task is simulated to be superimposed on the current load of the corresponding device, and the resource occupancy rate of each device after allocation is calculated. If the occupancy rate of a certain type of resource exceeds the preset threshold, or there is a timing conflict in the concurrent request of multiple tasks to the same resource, it is determined that the verification fails. At this time, the system will trigger the strategy adjustment mechanism: on the one hand, the overload devices are excluded from the candidate resource pool, and the similarity between the remaining devices and the task demand is recalculated; on the other hand, a load balancing factor is introduced to dynamically adjust the resource allocation weight. After adjustment, overload and conflict verification is performed again until a feasible allocation strategy is generated, ensuring the effectiveness of dynamic resource allocation and system stability.

[0056] Afterwards, task scheduling and allocation are continued: according to experimental requirements and equipment state information, an intelligent task scheduling algorithm based on the HONGBAO operating system is adopted to reasonably allocate experimental tasks to each device and ensure that each device can process in real time according to the task priority and cooperative relationship. The purpose of the task scheduling and allocation is to start from the task demand side, dynamically allocate tasks according to the device capability, and solve the problem of how to efficiently execute tasks. The specific implementation method is: first, collect all the experimental task information to be scheduled at present, construct a task description matrix , wherein represents the conversion requirement intensity of the jth experimental task on the ith type of resource, is the total number of tasks, is the number of resource types; collect the current running state parameters of each experimental device, and construct a device resource capability matrix , wherein represents the remaining amount of the ith type of resource that can be provided by the mth device, and M is the total number of devices; define the adaptation degree function of the task device: , wherein, is the priority weight of the ith type of resource, represents the resource matching efficiency of task j allocated to device m, and the match function is represented as the adaptation degree function between the task and the device; a bipartite graph of the task device is constructed, the edge weight is , and the task priority weight , the device load weight is introduced, and the optimization objective function is solved: , wherein , indicates whether the task is allocated to the device. Among them, the objective function needs to meet: each task is only allocated to one device: ; and the total resources of each device cannot be overloaded: . For solving the optimization objective function, the Hungarian algorithm is adopted for implementation.

[0057] Then, real-time data transmission is carried out: the low-delay communication protocol of the open-source HONGBAO system is used to ensure the data synchronization and streaming transmission between devices, to update the data state between devices in real time, and to reduce data delay and synchronization conflict. Specifically, in the actual running process, the system will allocate a dedicated data channel for each experimental task, and through the message queue mechanism and data cache optimization strategy, the data stream is collected and forwarded in real time. Each device marks the data state before sending data, and the receiving device judges the new and old and synchronization state of the data through the data identification and time stamp, so as to guarantee the consistency and order of the data.

[0058] Finally, experimental data analysis and report generation: after the completion of the experimental task, the system automatically summarizes and analyzes the device data, generates an experimental report based on the real-time task execution results, and displays the analysis results to the user through a visual interface, supporting the user to further optimize the experimental process and decision-making.

[0059] The above merely describes preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any person skilled in the art can modify or change the above disclosed technical content to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made to the above embodiments without departing from the technical solution of the present application, in accordance with the technical essence of the present application, still belongs to the protection scope of the technical solution of the present application.

Claims

1. A multi-device collaborative control method for a development laboratory kit based on an open-source Hongmeng system, characterized in that, Comprise the following steps: S1, device access and identity authentication: through the distributed device management platform in the Hongmeng operating system, access and identity authentication of multiple devices in the experiment box, and allocate a unique identifier for each device; S2, automatic discovery and dynamic connection of devices: automatically identify newly added or changed devices in the experiment box, and realize dynamic connection between devices; S3, device resource sharing: in the open source Hongmeng system, through a unified resource scheduling mechanism, realize the sharing and reasonable allocation of each device resource in the experiment box, including: resource information collection, scheduling strategy matching and dynamic resource allocation; Information collection: through the interface of open source Hongmeng system, collect the performance parameters and real-time load information of each device in the experiment box; Scheduling strategy matching: according to the collected performance parameters and real-time load information, combined with the demand of experimental task for resources, match the adaptive resource allocation strategy; Dynamic resource allocation: according to the matched strategy, through the unified scheduling of open source Hongmeng system, allocate the idle device resources to the experimental tasks with demand; S4, task scheduling and allocation: according to the experimental demand and device state information, use the intelligent task scheduling algorithm based on Hongmeng operating system to reasonably allocate the experimental tasks to each device, and ensure that each device can process in real time according to the task priority and cooperative relationship; S5, real-time data transmission: use the low delay communication protocol of open source Hongmeng system to ensure the data synchronization and streaming transmission between devices, update the data state between devices in real time, reduce data delay and synchronization conflict; S6, experimental data analysis and report generation: after the completion of experimental task, the system automatically summarizes and analyzes the device data, generates experimental report based on real-time task execution result, and shows the analysis result to user through visual interface, supports user to further optimize experimental process and decision; The specific implementation operation of the step S2, automatic discovery and dynamic connection of devices, is: Through the distributed software bus of the open-source HOMER system, multi-dimensional feature extraction is performed on the broadcast signals of the devices in the physical space of the test box to generate a space-time domain device fingerprint: wherein are respectively a time domain feature, a space feature, a radio frequency feature and a protocol feature, and m represents an index of the device. Constructing the dynamic topology graph of experiment box devices; Based on the device fingerprint similarity matrix S, the fingerprint similarity of devices m and n is calculated in the following manner: An adjacency matrix is dynamically generated according to the similarity threshold value, and if the similarity of two devices is greater than the threshold value, a connection relationship is established, that is, Wherein is the adjacency matrix, represents the similarity threshold value; Designing sliding event windows Temporal difference analysis of device states, detecting device changes: wherein is the state vector of device m at time t, and a change in the existing device state is assumed when the change flag satisfies the decision condition, and the change flag is determined as wherein is the change sensitivity parameter; When a new device joins or the state of an existing device changes, update the topology graph in real time. 2.The multi-device collaborative control method based on the open-source open-source monsoon system development experiment box of claim 1, wherein, The step S1 includes identification rule definition step, information collection step, authentication verification step and identifier allocation step for identity authentication and allocation of unique identifier for each device; Identification rule definition: according to the Hongmeng distributed device management specification, combined with the type and function attribute of experiment box devices, formulate the unique identifier coding rule, and clearly define the composition structure of device classification code, function code and serial number; Information collection: through the device access interface, collect the hardware characteristics, device type and manufacturer basic information of each device in the experiment box as the basis for authentication and identification; Authentication verification: use the security authentication framework of open source Hongmeng system to compare the collected device information with the preset device white list and security feature library, verify the legality and integrity of the device; Identifier allocation: for the authenticated devices, generate and allocate unique identifier according to the defined identification rule, associate the device basic information and authentication state, and synchronize the entry of identification database in the distributed device management platform.

3. The multi-device collaborative control method of the development experiment box based on the open-source microkernel system according to claim 2, characterized in that, The identifier dynamic verification and updating process is added after the identifier allocation step, including a verification trigger judgment step, a validity determination step, and an identifier updating step; Verification trigger judgment: set a timing verification period or trigger the identifier validity verification process when the device performs task type switching; Validity determination: extract the current hardware features and running state parameters of the device, compare them with the initial authentication information associated with the identifier, and determine whether the device has hardware tampering or identity impersonation risk; Identifier update: if there is a risk, re-execute the information collection and authentication verification steps to generate a new identifier to replace; If there is no risk, maintain the existing identifier and associated information.

4. The development experiment box multi-device collaborative control method based on the open source monsoon system according to claim 1, characterized in that, The step S3 scheduling strategy matching: according to the collected performance parameters and real-time load information, combined with the demand of experimental tasks for resources, the specific implementation steps of matching the adaptive resource allocation strategy are: The collected device performance parameters and real-time load information are used to construct a device resource state vector wherein, denotes the normalized parameter of the i-th type of resource. synchronizing parsing the experiment task's demand on resources, generating a task demand vector wherein is the demand weight on the i-th type of resource Similarity of a computing device resource vector to a task demand vector and filter out a device set with similarity greater than a set threshold as a candidate resource pool; Based on the candidate resource pool, the final allocation strategy is generated by weighted summation, and the calculation method is: Wherein, P represents the final allocation strategy, is the resource state vector of the mth candidate device, indicates the stability weight of the mth device performing the task, Wherein indicates the resource load rate of the mth device.

5. The development experiment box multi-device collaborative control method based on the open source monsoon system according to claim 1, characterized in that, Between S3 scheduling strategy matching and dynamic resource allocation, it is also necessary to check whether the candidate device will appear overload or resource conflict after allocating resources according to the strategy. If the verification fails, the strategy needs to be adjusted again.

6. The development experiment box multi-device collaborative control method based on an open source monsoon system according to claim 1, characterized in that, The step S4 task scheduling and allocation: according to the experimental demand and device state information, the specific implementation of using intelligent task scheduling algorithm based on the Hongmeng operating system to reasonably allocate experimental tasks to each device and ensure that each device can process in real time according to task priority and cooperative relationship is: First, collect all the information of the current experimental task to be scheduled, and construct the task description matrix where represents the conversion requirement intensity of the jth experimental task for the ith resource, is the total number of tasks, is the number of resource types; Collecting the current running state parameters of each experimental equipment to construct the equipment resource capability matrix wherein represents the remaining amount of the ith type of resource currently available for the mth equipment, and M is the total number of equipment Defining the fitness function of a task device: where, is the priority weight of the ith type of resource, denotes the resource matching efficiency of assigning task j to device m, the match function is denoted as the fitness function between the task and the device; A bipartite graph of the task devices is constructed, with edge weights and introducing task priority weights , device load weights , the optimization objective function is solved: where denotes whether a task is assigned to a device.

7. The development experiment box multi-device collaborative control method based on an open source monsoon system according to claim 6, characterized in that, The objective function needs to satisfy: each task is only assigned to one device: ; and the total resources of each device cannot be overloaded: .

Images