Event-triggered communication module state switching method and system

Abstract

The application provides a kind of communication module state switching method and system based on event triggering, it is related to communication technical field, first acquisition communication module in running multiple source event source information and carry out event influence level division;Then combine communication module current operating state with historical state switching database generation target associated network;Based on target associated network and historical stability record generation contains hardware parameter adjustment sequence etc. Dynamic path;Stability prediction is carried out to dynamic path;According to the result of pre-judgment, state switching execution instruction is generated and switching process data is generated;Finally, based on switching process data, update historical state switching database and target associated network form closed loop optimization, so as to realize the dynamic optimization of communication module state switching, improve the adaptability and stability of switching.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to a communication module state switching method and system based on event triggering. BACKGROUND

[0002] In the context of continuous evolution of communication technology, communication modules, as the core components for realizing communication functions, are widely used in various intelligent terminal devices. In actual operation, communication modules need to be flexibly switched according to different business scenarios and device states to ensure the efficiency and stability of communication.

[0003] However, the existing communication module state switching method has many deficiencies. On the one hand, traditional methods often only rely on single event source information to trigger state switching, such as switching only according to software instructions or hardware signals, ignoring the complex correlation and influence level between multiple event sources. This may lead to inaccurate judgment of the timing and necessity of state switching when facing complex and variable communication environments, resulting in unnecessary switching operations, increasing system overhead and power consumption, and even affecting communication quality.

[0004] On the other hand, the existing state switching path is usually a pre-set fixed mode, lacking the ability to dynamically adjust according to real-time events and historical data. In the case of changing communication scenarios, the fixed switching path may not adapt to new needs, leading to decreased system stability after switching, communication interruption, data loss and other problems. In addition, the lack of effective prediction and evaluation mechanism for state switching stability makes the switching operation somewhat blind, making it difficult to guarantee the reliability and safety of the switching process. SUMMARY

[0005] In view of the above-mentioned problems, in combination with the first aspect of the present application, the embodiments of the present application provide a communication module state switching method based on event triggering, which comprises:

[0006] Collecting multi-source event source information in the running process of the communication module, performing event influence level division processing on the multi-source event source information to obtain event layering results, the multi-source event source information including hardware running abnormal signal, software instruction interaction information and external device association request information;

[0007] Obtaining the current running state of the communication module and the historical state switching database, combining the event layering results with the current running state of the communication module to generate a target association network, the historical state switching database storing historical event types, corresponding switching paths and post-switching stability records;

[0008] Based on the target associated network and the stability record in the historical state switching database, a dynamic path of communication module state switching is generated, the dynamic path includes a hardware parameter adjustment sequence, a software process scheduling sequence and an external interaction adaptation strategy;

[0009] The dynamic path is subjected to state switching stability prediction processing, and a prediction result is generated in combination with a current hardware load threshold and a software dependency relationship of the communication module;

[0010] According to the prediction result, the dynamic path is converted into a state switching execution instruction, and a switching operation is executed by inputting the communication module to generate switching process data;

[0011] Based on the stability cycle estimation in the switching process data and the prediction result, the historical state switching database and the target associated network are updated to form a state switching closed-loop optimization mark.

[0012] In another aspect, the embodiment of the present application also provides a communication module state switching system based on event triggering, characterized in that it comprises:

[0013] A processor; a machine readable storage medium for storing machine executable instructions of the processor; wherein the processor is configured to execute the machine executable instructions to perform the above-mentioned communication module state switching method based on event triggering.

[0014] In another aspect, the embodiment of the present application also provides a computer program product, the computer program product comprising machine executable instructions stored in a computer readable storage medium, a processor of a computer device reading the machine executable instructions from the computer readable storage medium, and the processor executing the machine executable instructions, so that the computer device executes the above-mentioned communication module state switching method based on event triggering.

[0015] Based on the above aspects, by collecting multi-source event source information in the running process of the communication module and performing event influence level division processing, combining the current running state and the historical state switching database to generate a target associated network, the internal relationship between events and states can be deeply mined, and the state switching decision is more in line with the actual running situation. Based on the target associated network and the stability record in the historical state switching database, a dynamic path is generated, the state switching path is dynamically optimized, the switching strategy can be automatically adjusted according to real-time events and historical experience, and the adaptability and flexibility of state switching are improved. The state switching stability of the dynamic path is predicted and processed, the current hardware load threshold and software dependency of the communication module are fully considered, the risk of unstable situation in the switching process is effectively reduced, and the reliability of the switching operation is enhanced. According to the prediction result, a state switching execution instruction is generated, and switching process data is generated, and finally the historical state switching database and the target associated network are updated based on the switching process data, forming a closed-loop optimization mechanism for state switching, which can continuously accumulate experience, self-improve and continuously improve the performance and stability of the communication module state switching. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 is the execution flow diagram of the communication module state switching method based on event triggering provided by the embodiment of the application.

[0017] Figure 2 is a schematic diagram of exemplary hardware and software components of the communication module state switching system based on event triggering provided by the embodiment of the application. DETAILED DESCRIPTION

[0018] The application will be specifically described below with reference to the drawings of the specification, Figure 1 is a flow diagram of the communication module state switching method based on event triggering provided by an embodiment of the application, and the communication module state switching method based on event triggering will be described in detail below.

[0019] Step S110: Collecting multi-source event source information in the running process of the communication module, performing event influence level division processing on the multi-source event source information, obtaining event layering results, and the multi-source event source information including hardware running abnormal signal, software instruction interaction information and external device associated request information.

[0020] In this embodiment, the communication module is integrated in a user-side intelligent gateway, which is connected to the user's smart phone, smart TV, smart lamp and security camera through wired Ethernet, Wi-Fi, Bluetooth and other interfaces. When collecting multi-source event source information, three types of information are obtained through hardware monitoring chips, software log analysis modules and external interaction protocol analyzers, and event layering is completed through an influence coefficient calculation model and a level determination rule.

[0021] Step S111: Collecting hardware operation anomaly signals from the hardware monitoring port of the communication module, the hardware operation anomaly signals including power module voltage fluctuation, transmission module signal attenuation and interface module connection frequency change.

[0022] The hardware monitoring port of the communication module includes the ADC sampling port of the power module, the signal strength detection port of the transmission module and the state monitoring pin of the interface module. The ADC sampling port of the power module adopts a 12-bit successive approximation ADC, collects the voltage at the output end of the power supply through a voltage dividing resistor network, and the sampling frequency is matched with the output ripple period of the power module. When the sampling value deviates from the reference range for three consecutive times, a voltage fluctuation signal is generated, which includes the sampling timestamp, the current voltage value and the deviation direction information.

[0023] The signal strength detection port of the transmission module is integrated in the RF front-end chip, and signal attenuation monitoring is achieved by detecting the RSSI value of the received signal. When the RSSI value continuously decreases and the decrease amplitude exceeds the chip's preset sensitivity threshold, the attenuation start time, attenuation rate and current RSSI value are recorded to form signal attenuation data.

[0024] The state monitoring pin of the interface module is connected to the connection detection circuit of each physical interface, and the connection establishment and disconnection times per unit time are counted by detecting the change of the pin level. For Ethernet interfaces, the monitoring pin detects link pulse signals; for Wi-Fi interfaces, the monitoring pin detects the trigger signals of association request and disassociation frames, and finally integrates to form connection frequency change data, which is stored by interface type.

[0025] Step S112: Extracting software instruction interaction information from the software log of the communication module, the software instruction interaction information including instruction sending interval, instruction execution delay and instruction retry count.

[0026] The software log of the communication module adopts a circular storage mechanism and is stored in the log partition of the on-chip Flash. The log record format complies with the syslog protocol, and each log entry includes the timestamp, process ID, instruction type, instruction parameter and execution status fields. When extracting software instruction interaction information, first filter the instruction interaction logs in the latest monitoring period by timestamp through the log parsing module, and then process them by instruction type.

[0027] For instruction sending interval, for the same type of instructions sent by the same process, extract the sending timestamp of each instruction, calculate the time difference between adjacent two instructions, and form the sending interval sequence of the same type of instructions. For instruction execution delay, extract the instruction sending timestamp and execution completion status feedback timestamp, calculate the difference between them, and if there is an interruption during instruction execution, the interruption duration needs to be included in the execution delay.

[0028] For command retrieval counts, retrieve command records marked "retry" from the log, group them by command ID to count the number of retries, and record the retry reason field, such as "timeout without response" or "verification error", to ensure that the number of retries is associated with the specific cause of the failure.

[0029] Step S113: Collect external device association request information from the external interaction port of the communication module. The external device association request information includes the type of the request initiating device, the amount of request data, and the timeliness requirement for request response.

[0030] The communication module's external interaction ports include an Ethernet RJ45 port, a Wi-Fi 6 wireless port, and a Bluetooth 5.0 port, each equipped with a protocol parsing unit. When an external device initiates an association request, the protocol parsing unit first parses the frame header field of the request frame. For Ethernet frames, the device type is identified by the source MAC address prefix; for Wi-Fi frames, the device type is identified by the device manufacturer information field in the ProbeRequest frame; and for Bluetooth frames, the device type is identified by the device category field in the broadcast packet.

[0031] The requested data volume is obtained by parsing the payload length field of the request frame. For TCP requests, the actual amount of data to be transmitted needs to be determined by combining the window size field in the packet header; for UDP requests, the datagram length field is extracted directly. The required response time is determined by parsing the service type field in the request frame. For example, a real-time preview request for a smart camera corresponds to an immediate response, a status query request for smart lights corresponds to a regular response, and a firmware update request for a smart TV corresponds to a delayed response.

[0032] Step S114: Analyze the impact range of the hardware operation anomaly signal, determine the number of hardware components affected by the signal and the importance of the components, and generate a hardware impact coefficient.

[0033] First, the components affected by each abnormal signal are identified using a hardware correlation diagram. This diagram predefines the connection relationships between the power supply module, transmission module, interface module, and components such as the core processor and memory. If the power supply module voltage fluctuates, the diagram shows the affected power supply unit of the core processor, the power management chip of the memory, and the power supply circuits of each interface module, and counts the number of these components.

[0034] The importance of the components is divided into three levels according to the function criticality, the core processor and the radio frequency transmission chip are level one, the memory and the power management chip are level two, and the interface physical layer chip is level three, and each level corresponds to a preset weight coefficient. The number of affected components is standardized according to the total number of components, and then weighted summation is performed with the weight coefficients of the components at each level, and the summation result is normalized to map to the interval of 0 to 1 to obtain a hardware influence coefficient. The larger the hardware influence coefficient is, the more significant the influence of the hardware abnormality is.

[0035] Step S115: performing influence analysis on the software instruction interaction information to determine the interruption range of the instruction abnormality to the software process chain and the process recovery difficulty, and generating a software influence coefficient.

[0036] First, the instruction abnormality determination module identifies abnormal instructions. When the instruction execution delay exceeds the preset threshold corresponding to the instruction type, or the retry number reaches the maximum retry upper limit set by the process, it is determined that the instruction is abnormal. For abnormal instructions, the process dependency graph is used to trace the process chain to which the abnormal instructions belong. The process dependency graph stores the calling relationship between processes in the form of a directed graph, the nodes are processes, and the edges are calling relationships. The attributes of the edges mark the calling frequency and data interaction amount.

[0037] The interruption range is determined according to the process chain structure. If the abnormal instruction belongs to an upstream process in the link, the interruption range covers the process and all downstream called processes; if it is a downstream process, the interruption range only includes the process and the upstream process that directly calls it. The process recovery difficulty is evaluated according to the startup dependent resources, initialization time and data recovery mechanism of the process. The fewer the dependent resources, the shorter the initialization time, and the more the processes with data backup mechanism, the lower the recovery difficulty. The recovery difficulty corresponds to a preset weight value.

[0038] The number of processes corresponding to the interruption range is standardized, and the software influence coefficient is generated after weighted summation and normalization of the recovery difficulty weight value. The numerical range is 0 to 1.

[0039] Step S116: performing association influence analysis on the external device association request information to determine the influence breadth of the request interruption to the external device interaction link and the link reconstruction cost, and generating an external influence coefficient.

[0040] The influence breadth is determined based on the user-side device interaction topology graph, which is dynamically constructed by the device discovery module of the communication module and records the connection relationship between each external device and the communication module and the indirect interaction relationship between devices. When the association request of a certain external device is interrupted, the topology graph displays other devices directly connected to the device and downstream devices dependent on the data of the device, and the number of interaction links corresponding to these devices is counted, which is the influence breadth.

[0041] The link reconstruction cost is evaluated from time cost and resource cost. The time cost includes link renegotiation time and protocol handshake time. The resource cost includes CPU resource, memory resource and bandwidth resource consumed in the reconstruction process. According to the cost evaluation result, the influence breadth is standardized, and the cost weight is weighted and summed, and then normalized to obtain the external influence coefficient, the numerical range is 0 to 1.

[0042] Step S117: The hardware influence coefficient, the software influence coefficient and the external influence coefficient are associated and integrated, and the influence level division standard preset by the communication module is referred to, to generate an event hierarchical result. The event hierarchical result includes an event core influence level, an associated influence level and an indirect influence level, and different influence levels correspond to different state switching response priorities.

[0043] The influence coefficient association module binds the hardware influence coefficient, the software influence coefficient and the external influence coefficient according to the event occurrence timestamp, to ensure that the three correspond to the same event source. The preset influence level division standard is stored in a configuration file. The core influence level requires that the hardware influence coefficient, the software influence coefficient and the external influence coefficient are all greater than a first threshold value, and at least one coefficient is greater than the product of the first threshold value and a correction coefficient. The associated influence level requires that two coefficients are greater than a second threshold value, and a third coefficient is greater than a third threshold value. The indirect influence level requires that at least one coefficient is greater than a fourth threshold value, and the first two levels are not reached. The first threshold value> the second threshold value> the third threshold value> the fourth threshold value, and the correction coefficient is an empirical value less than 1.

[0044] After comparing the integrated coefficients with the standard, the event level is determined, the core influence level is assigned the highest response priority, the associated influence level is assigned a medium priority, and the indirect influence level is assigned the lowest priority, to finally form an event hierarchical result including event ID, influence level, response priority and three influence coefficients.

[0045] Step S120: Obtain the current running state of the communication module and the historical state switching database, combine the event hierarchical result and the current running state of the communication module, and generate a target associated network. The historical state switching database stores historical event types, corresponding switching paths and stability records after switching.

[0046] The state acquisition module acquires the running state parameters of the communication module in real time, retrieves the historical state switching data from the local database or the cloud storage, and then associates and maps the level information in the event hierarchical result with the current state elements and the historical data through the network construction module, to construct a target associated network including nodes and edges, and intuitively present the influence and mapping relationship between elements.

[0047] Step S121: obtaining a current running state of the communication module, the current running state of the communication module including a hardware component running parameter, a software process active state, and an external interaction link state.

[0048] In this embodiment, the hardware component running parameter is collected by a distributed monitoring unit, and the running parameter of the core processor includes a current working frequency, a core temperature, and a cache hit rate. The frequency is obtained by reading a PLL configuration register of the processor, the temperature is collected by a temperature sensor built in the processor, and the cache hit rate is counted by a performance monitoring unit of the processor.

[0049] The running parameter of the memory includes an SRAM usage rate, a Flash erasing and writing number, and a data read and write rate. The SRAM usage rate is obtained by calculating a ratio of an allocated memory to a total memory, the Flash erasing and writing number is stored in a state register of the Flash, and the read and write rate is calculated by counting a read and write data amount in a unit time.

[0050] The running parameter of each interface module includes a link rate and a frame error rate of an Ethernet interface, a channel occupancy rate and a retransmission rate of a Wi-Fi interface, and a connection number and a frequency hopping interval of a Bluetooth interface. These parameters are read by a state register of an interface controller.

[0051] The software process active state is obtained by a process management unit. The process management unit periodically traverses a process control block linked list, and extracts a process name, a process ID, a current state (running, ready, blocking, and termination), an occupied CPU time, a memory usage, and a number of opened file descriptors of each process.

[0052] For the process in the blocking state, a blocking reason needs to be recorded, such as waiting for I / O completion, waiting for a semaphore, or waiting for a timer timeout. Meanwhile, a pipe, a message queue, and a shared memory interaction between processes are recorded by an inter-process communication monitoring module, to determine a dependency relationship between processes.

[0053] The external interaction link state is collected by a link monitoring module. The link monitoring module listens to a protocol data unit of each external interaction port, extracts a connection state of the link, a used communication protocol version, a data transmission error code rate, a round-trip delay, and a throughput.

[0054] For the established link, the connection state is monitored by sending a link probe frame. For the link in a connection establishment process, an interaction progress of a handshake frame is monitored. The error code rate is calculated by counting a ratio of an error bit number to a total bit number in received data. The round-trip delay is obtained by recording a time difference between sending and receiving a response of a probe frame. The throughput is determined by counting a total sum of successfully transmitted data packet sizes in a unit time.

[0055] Step S122: retrieve the historical state switching database, which contains historical event records, historical switching path records, and historical post-switching stability records, the historical event records correspond to the event types in the event hierarchical result, the historical switching path records contain historical hardware parameter adjustment sequences, historical software process scheduling orders, and historical external interaction adaptation strategies.

[0056] The historical state switching database adopts a SQLite database and is deployed on an external SD card of the communication module. The database contains three core tables: an event record table, a switching path table, and a stability record table. When the database is retrieved, a connection is established through a database driver program, and a query statement is executed to obtain data.

[0057] The event record table stores the event ID, event type code, occurrence time, hardware influence coefficient, software influence coefficient, external influence coefficient, and influence level of historical events. The event type code corresponds to the event type in the event hierarchical result through a preset encoding mapping table.

[0058] The switching path table is associated with the event record table through the event ID. Each record contains a hardware parameter adjustment sequence (parameter name, adjustment order, pre-adjustment value, post-adjustment value), a software process scheduling order (process ID, operation type, operation order, resource allocation ratio), and an external interaction adaptation strategy (link ID, protocol adjustment content, cache configuration parameter).

[0059] The stability record table is also associated through the event ID and contains the switching completion time, post-switching stable running time, number of exceptions during the period, exception type, and exception recovery method.

[0060] Step S123: associate the event core influence level in the event hierarchical result with the hardware component running parameters in the current running state of the communication module, identify the hardware sensitive parameters corresponding to the core influence level, and the hardware sensitive parameters are the most significant parameters in the hardware component running that respond to events.

[0061] First, extract the event ID, event type, and hardware influence coefficient of the core influence level from the event hierarchical result. Then, according to the event type, filter the hardware component running parameters related to this type of event in the current running state of the communication module, such as power supply module parameters for power supply events and transmission module parameters for transmission events.

[0062] The correlation analysis module calculates the Pearson correlation coefficient of each screened parameter and the hardware influence coefficient, and the parameters with an absolute value of the correlation coefficient greater than a preset threshold are initially listed as candidate sensitive parameters. Then, the parameter sensitivity test module performs perturbation test on the candidate parameters, that is, slightly adjusts the parameter value, observes the change amplitude of the hardware influence coefficient, and the parameter with the largest change amplitude is determined as the hardware sensitive parameter, for example, the output voltage adjustment coefficient of the tested power module is the hardware sensitive parameter for the power voltage fluctuation event.

[0063] Step S124: associate the associated influence level in the event layering result with the software process active state in the current running state of the communication module, identify the software dependent process corresponding to the associated influence level, and the software dependent process is a process that needs to be scheduled preferentially in the software system affected by the event.

[0064] Extract the event type and software influence coefficient of the associated influence level in the event layering result, combine the current software process active state of the communication module, and determine the initial process affected by the event through the process influence analysis module. The initial process is usually a process directly related to the event type, such as an instruction parsing process corresponding to an instruction execution delay event.

[0065] Based on the process dependency relationship recorded in the software process active state, a process dependency tree with the initial process as the root node is constructed, and the child nodes of the tree are downstream processes dependent on the initial process. The dependency strength between each child node and the root node is calculated by the dependency strength evaluation module, and the dependency strength is determined according to the interaction frequency, data interaction amount and interaction necessity between processes.

[0066] Processes with a dependency strength greater than a preset threshold are listed as software dependent processes, and the priority of the dependent processes is determined according to the dependency strength sorting, that is, the higher the priority, the more need to be processed preferentially in subsequent scheduling.

[0067] Step S125: associate the indirect influence level in the event layering result with the external interaction link state in the current running state of the communication module, identify the external link node corresponding to the indirect influence level, and the external link node is a node in the external interaction that needs to adjust the adaptation strategy affected by the event.

[0068] Extract the event features and external influence coefficients of the indirect influence level from the event layering result, match the current external interaction link state of the communication module according to the event features, and determine the external interaction link that may be affected, for example, a wireless interface connection frequency abnormal event matches a Wi-Fi link.

[0069] Each external interaction link is divided into a source node (communication module interface), an intermediate node (such as a wireless router, if present), and a destination node (external device) according to node types. The importance of each node in the link is evaluated by a node importance evaluation module. The importance is determined according to the degree of influence of the node on the connectivity of the link, the amount of data forwarding, and the difficulty of fault recovery.

[0070] Nodes with importance greater than a preset threshold are identified as external link nodes, and the type, IP address or MAC address, and current running state of the node are recorded.

[0071] Step S126: Map and associate the hardware sensitive parameters, software dependent processes, and external link nodes with the historical event records and historical switching path records in the historical state switching database, present the corresponding relationship between event layers, current state elements, and historical switching data in the form of a graph, and generate a target associated network. The target associated network includes nodes and associated edges. The nodes represent event levels, state elements, and historical switching paths, and the associated edges represent the influence relationship and mapping relationship between the nodes.

[0072] The target associated network adopts a directed graph structure and is constructed by a graph theory modeling tool. First, define the node types and attributes: the event level node includes node ID, level type, and event ID attributes; the state element node includes node ID, element type, and element identifier attributes, where the element type is divided into hardware sensitive parameters, software dependent processes, and external link nodes; the historical data node includes node ID, data type, and associated event ID attributes, and the data type is divided into historical events and historical switching paths.

[0073] Establish the associated edges between nodes: the event level node points to the corresponding state element node, and the edge attribute is the influence degree (determined by the influence coefficient); the state element node points to the historical data node, and the edge attribute is the mapping similarity (calculated by the matching degree of the current element and the historical data); the historical event node points to the corresponding historical switching path node, and the edge attribute is the corresponding weight.

[0074] The weight of the associated edge is determined by the standardized influence degree, mapping similarity, and corresponding weight. The greater the weight value, the closer the relationship between the nodes. The constructed directed graph is stored as a GraphML format file, which can be loaded and viewed by a network visualization tool to query the nodes and edges, and support path analysis.

[0075] Step S130: Based on the target associated network and the stability records in the historical state switching database, generate a dynamic path for communication module state switching. The dynamic path includes a hardware parameter adjustment sequence, a software process scheduling order, and an external interaction adaptation strategy.

[0076] The node correlation relationship in the target correlation network is mined by the network analysis module, the effect difference of different adjustment schemes is analyzed in combination with the historical stability record, and a dynamic path suitable for the current event and running state is generated through standardized scoring, scheme screening and integration optimization.

[0077] Step S131: Extract the hardware sensitive parameters corresponding to the event core influence level from the target correlation network, combine the historical hardware parameter adjustment sequence under the same hardware sensitive parameter in the historical state switching database, analyze the first influence difference of different adjustment sequences on hardware stability, and the first influence difference is reflected by the hardware stability time length in the historical switching stability record.

[0078] Through the node query interface of the target correlation network, the hardware sensitive parameter nodes corresponding to the node ID of the event core influence level are extracted, and the parameter name and current value are obtained. With the parameter name as the query condition, all historical hardware parameter adjustment sequences corresponding to the same parameter are retrieved in the switching path table of the historical state switching database, and each sequence contains the adjustment steps, adjustment amplitude and adjustment order of the parameter.

[0079] For each historical adjustment sequence, the stability record table in the historical state switching database is associated with the event ID to extract the hardware stability time length corresponding to the sequence. All historical adjustment sequences are sorted by hardware stability time length from long to short, and the difference in stability time length between adjacent sequences is calculated to quantify the first influence difference of different adjustment sequences on hardware stability. At the same time, the hardware component running parameter changes corresponding to each sequence are recorded, such as voltage values before and after adjustment, signal strength, etc.

[0080] Step S132: Extract the software dependent processes corresponding to the correlation influence level from the target correlation network, combine the historical software process scheduling order under the same software dependent process in the historical state switching database, analyze the second influence difference of different scheduling orders on software response speed, and the second influence difference is reflected by the software instruction execution delay in the historical switching stability record.

[0081] Through the node traversal function of the target correlation network, the software dependent process node set corresponding to the correlation influence level node is located, and the process ID and function description of each software dependent process are obtained. The corresponding historical software process scheduling order is queried in the switching path table of the historical state switching database according to the process ID, and each scheduling order contains the start order, close order and resource allocation ratio of the process.

[0082] For each historical scheduling sequence, the software instruction execution delay data is extracted by associating the stability record table with the event ID. The average execution delay of different scheduling sequences under the same software dependent process is calculated by grouping and counting the above delay data, and the difference in average execution delay is taken as the second quantitative indicator of the difference in influence. In addition, it is also necessary to analyze the correlation between process operation logic in the scheduling sequence and software instruction execution delay, such as whether starting the core dependent process first can significantly reduce the execution delay.

[0083] Step S133: Extract the external link node corresponding to the indirect influence level from the target associated network, combine the historical external interaction adaptation strategy under the same external link node in the historical state switching database, analyze the third influence difference of different adaptation strategies on link stability, and the third influence difference is reflected by the link interruption frequency in the historical switching stability record.

[0084] Using the node screening function of the target associated network, the corresponding external link node is extracted according to the event characteristics of the indirect influence level, the MAC address, the connection protocol type and the link identifier of the node are obtained. Taking the link identifier as the index, the historical external interaction adaptation strategy corresponding to the same external link node is searched in the switching path table of the historical state switching database, which includes protocol adjustment scheme, cache configuration parameter and link monitoring frequency, etc.

[0085] For each historical adaptation strategy, the link interruption frequency data is extracted by associating the stability record table with the event ID. The average interruption frequency under different adaptation strategies is counted, the difference between the frequencies is calculated to determine the third influence difference. At the same time, the changes of link throughput, bit error rate and other parameters after the adjustment of the adaptation strategy are recorded, and the corresponding relationship between the adaptation strategy and the link stability index is analyzed.

[0086] Step S134: Standardize the first influence difference, the second influence difference and the third influence difference, convert them into unified relative influence scores, and select the adapted hardware parameter adjustment direction, software process scheduling priority and external interaction adaptation mode for the current event based on the relative influence scores.

[0087] First, determine the original value range of the first influence difference, the second influence difference and the third influence difference. For the first influence difference (hardware stability duration difference), the difference between the maximum value and the minimum value is taken as the denominator, and the difference between a certain difference value and the minimum value is taken as the numerator, to calculate the standardized value of the difference. For the second influence difference (software instruction execution delay difference) and the third influence difference (link interruption frequency difference), the matching method is adopted, but the difference value is taken after being inverted, and then standardized, to ensure that the larger the standardized value is, the better the influence is.

[0088] The standardized values of the three dimensions are multiplied by preset weight coefficients, which are set according to the priority of each influence level in the event stratification result. The first influence difference corresponding to the core influence level has the highest weight, the second influence difference corresponding to the associated influence level is second, and the third influence difference corresponding to the indirect influence level is the lowest. The relative influence score of each historical adjustment scheme is obtained by weighted summation, and the hardware parameter adjustment direction, software process scheduling priority and external interaction adaptation mode corresponding to the scheme with the highest score are selected as the adaptation scheme of the current event.

[0089] Step S135: converting the hardware parameter adjustment direction into a specific hardware parameter adjustment sequence, including the order and adjustment amplitude gradient of parameter adjustment.

[0090] According to the selected hardware parameter adjustment direction, the list of hardware sensitive parameters that need to be adjusted is sorted according to the influence degree of the parameters on the stability of the hardware, and the parameters with high influence degree are adjusted first. For example, if the hardware sensitive parameters are the output voltage adjustment coefficient of the power module and the signal gain of the transmission module, the output voltage adjustment coefficient has a greater impact on the overall hardware power supply stability and should be adjusted first.

[0091] For each parameter, set the adjustment amplitude gradient. The initial adjustment amplitude is referenced from the optimal amplitude value in the historical adjustment sequence, and if the hardware stability duration after adjustment does not meet the expectation, the amplitude is gradually increased or decreased according to the preset gradient. At the same time, the adjustment interval time of each parameter is determined to ensure that the next parameter adjustment is performed after the previous parameter adjustment is completed and stable for a period of time, avoiding parameter adjustment conflicts. The final hardware parameter adjustment sequence needs to include parameter name, adjustment order, initial adjustment amplitude, gradient change rule and adjustment interval time.

[0092] Step S136: converting the software process scheduling priority into a specific software process scheduling sequence, including the order of process shutdown and startup logic and process resource allocation ratio.

[0093] According to the software process scheduling priority, the order of process operation is determined. When the process is closed, first close the non-dependent process with low association degree to the current event, then close the secondary software dependent process, and finally close the main software dependent process, to avoid system abnormalities caused by premature shutdown of the core process. When starting the process, follow the reverse order, first start the main software dependent process, and then start the secondary software dependent process and the non-dependent process after the initialization is completed.

[0094] According to the process priority allocation resource ratio, the main software dependent process allocates the highest proportion of CPU resources and memory space, the secondary software dependent process allocates medium proportion of resources, and the non-dependent process allocates basic guarantee resources. At the same time, the resource dynamic adjustment mechanism is set for the process. When the instruction execution delay of a process exceeds the threshold value, the resource allocation ratio is temporarily increased, and after the delay returns to normal, it is called back to the original ratio, forming a software process scheduling sequence containing process operation logic, resource allocation benchmark ratio and dynamic adjustment rule.

[0095] Step S137: converting the external interaction adaptation mode into a specific external interaction adaptation strategy, including interaction protocol adjustment mode and data transmission caching mechanism.

[0096] For the selected external interaction adaptation mode, determine the interaction protocol adjustment mode. If the link interruption frequency of the current protocol is high, switch to a protocol version with stronger compatibility, or adjust the transmission window size, retransmission timeout time and other parameters in the protocol. For example, when the Wi-Fi link is frequently interrupted due to retransmission timeout, increase the retransmission timeout parameter to reduce the retransmission trigger frequency.

[0097] Design a data transmission caching mechanism to set the size of the cache area and the survival time of the cached data according to the request data volume and response timeliness requirements of the external device. For real-time response type requests, set a small cache area and a short survival time to ensure data real-time performance; for delay response type requests, set a large cache area and a long survival time to avoid data loss. At the same time, establish a priority scheduling mechanism for cached data to preferentially transmit cached data with high timeliness requirements, forming an external interaction adaptation strategy containing protocol adjustment parameters, cache configuration information and data scheduling rules.

[0098] Step S138: integrate the hardware parameter adjustment sequence, software process scheduling sequence and external interaction adaptation strategy to generate a dynamic path for communication module state switching. Each link in the dynamic path is provided with an associated identifier to reflect the cooperative relationship between hardware, software and external interaction adjustment.

[0099] Integrate the hardware parameter adjustment sequence, software process scheduling sequence and external interaction adaptation strategy according to the time axis and logical dependency. When the hardware parameter adjustment reaches the preset amplitude, trigger the software process scheduling operation corresponding to the associated identifier, for example, after the power module voltage is adjusted to a stable value, trigger the start operation of the instruction analysis process. When the software process completes initialization and enters a stable running state, trigger the execution of the external interaction adaptation strategy corresponding to the associated identifier, for example, after the software dependent process is started, the protocol adjustment operation of the external link is started.

[0100] A timestamp, an associated link ID and a state identifier are added to each link in the dynamic path, the timestamp records the scheduled execution time of the link, the associated link ID identifies other links triggered or triggered by the link, and the state identifier records the execution state of the link (to be executed, in execution, completed, and abnormal). The integrated content is stored as a path file in XML format, including three sub-nodes of hardware adjustment, software scheduling and external adaptation, and each sub-node lists specific links and associated information in execution order to form a complete dynamic path.

[0101] Step S140: State switching stability prediction processing is performed on the dynamic path, a prediction result is generated in combination with the current hardware load threshold and software dependency relationship of the communication module.

[0102] The dynamic path simulation execution module simulates the running of each link in the dynamic path, analyzes possible risks such as load overrun and dependency rupture in combination with the current hardware load capacity and software dependency structure, quantifies the risk level and estimates the stable period, and finally forms a prediction result including path adaptation degree and stable period estimation.

[0103] Step S141: Key adjustment nodes in the hardware parameter adjustment sequence are extracted from the dynamic path, the key adjustment nodes are nodes with an adjustment amplitude greater than a set amplitude or affecting multiple hardware components.

[0104] The hardware parameter adjustment sequence in the dynamic path is traversed, and nodes with an adjustment amplitude exceeding a preset threshold are filtered out, the adjustment amplitudes of these nodes are significantly higher than those of other conventional adjustment nodes. At the same time, nodes that will affect the running parameters of two or more hardware components after adjustment are identified, for example, adjusting the main power supply voltage of the power supply module will simultaneously affect the power supply state of the core processor and the transmission module. These two types of nodes are marked as key adjustment nodes, and the parameter name, adjustment amplitude, affected hardware components and planned adjustment time corresponding to the nodes are recorded.

[0105] Step S142: The current hardware load threshold of the communication module is obtained, the current hardware load threshold of the communication module includes the maximum bearing voltage, the maximum transmission rate and the maximum heat dissipation capacity of each hardware component.

[0106] The hardware load monitoring module collects the load threshold parameters of each hardware component, the maximum bearing voltage of the core processor is read from its specification sheet, the maximum transmission rate is determined by testing its data processing capacity, and the maximum heat dissipation capacity is calculated according to the fin specifications, fan speed and environmental temperature.

[0107] The maximum bearing voltage of the power module is determined by the design parameters of its voltage stabilizing circuit, and the maximum output current is determined by the limiting parameters of the power management chip. The maximum transmission rate of the transmission module refers to the performance indicators of its radio frequency chip, and the maximum signal transmission power is limited by the regulatory certification parameters. The maximum transmission rate of the interface module is determined according to the standards of the interface type (such as Ethernet, Wi-Fi), and the maximum number of connections is calculated by the resource limitation of the interface controller. The above parameters are classified and arranged according to hardware components to form a current hardware load threshold list.

[0108] Step S143: Simulate the hardware parameter adjustment process corresponding to the key adjustment node, calculate the actual load values of each hardware component after adjustment, compare the actual load values with the current hardware load threshold of the communication module, and analyze the hardware load overrun risk.

[0109] Step S1431: Establish a load calculation model for each hardware parameter adjustment item corresponding to the key adjustment node, which is based on the parameter-load correlation of hardware components and is used to reflect the corresponding relationship between parameter adjustment amount and load change amount.

[0110] For each hardware parameter adjustment item, collect the corresponding data of parameter adjustment amount and load change amount of each hardware component in the historical adjustment process, establish a parameter-load correlation equation through regression analysis, and form a load calculation model. For example, for the power voltage adjustment item, the model input is the voltage adjustment amount, and the output is the current load change amount of the core processor, transmission module and other components; for the signal gain adjustment item, the model input is the gain adjustment amount, and the output is the power load change amount of the transmission module.

[0111] The coupling relationship between hardware components needs to be included in the model, for example, the load change of a component will indirectly affect the load of other associated components. By introducing a coupling coefficient to correct the load calculation result, the accuracy of the model is ensured.

[0112] Step S1432: Input the adjustment amplitude and adjustment duration in the hardware parameter adjustment item into the load calculation model, and output the real-time load change curve of each hardware component in the adjustment process.

[0113] The adjustment amplitude, adjustment start time and adjustment duration of the hardware parameter adjustment item in the key adjustment node are input into the load calculation model, and the model calculates the load values of each hardware component according to the preset time step (such as every millisecond), and connects the load values at different time points to form a real-time load change curve. The curve needs to mark the load rising stage, stable stage (if there is a stable period in the adjustment process) and final stable load value, and intuitively present the dynamic change of the load in the adjustment process.

[0114] Step S1433: Extract the peak load value in the real-time load change curve as the actual load value of each hardware component after adjustment.

[0115] Traverse all data points of the real-time load change curve to determine the maximum load value of each hardware component during the adjustment process, i.e., the peak load value. If the load rises first and then falls and stabilizes during the adjustment process, the peak load value is the maximum value in the rising phase. If the load continues to rise and stabilizes at a certain value after adjustment, the peak load value is the load value after stabilization. Record the peak load value of each hardware component as the actual load value after adjustment.

[0116] Step S1434: Compare the actual load value by hardware component type with the corresponding current hardware load threshold of the communication module one by one.

[0117] Classify the actual load value by core processor, power module, transmission module, interface module, etc. and compare it with the corresponding parameters in the current hardware load threshold one by one. For example, compare the actual current load value of the core processor with the maximum bearing current threshold, compare the actual power load value of the transmission module with the maximum output power threshold, and record the results of each comparison (not exceeding the limit, slightly exceeding the limit, and seriously exceeding the limit).

[0118] Step S1435: Count the number of hardware components whose actual load value exceeds the current hardware load threshold of the communication module and the exceeding limit amplitude, which is the difference between the actual load value and the current hardware load threshold of the communication module.

[0119] Statistically analyze the comparison results to calculate the number of hardware components whose actual load value exceeds the threshold, and distinguish the number of components that are slightly exceeding the limit (exceeding limit amplitude less than 10% of the threshold) and seriously exceeding the limit (exceeding limit amplitude greater than or equal to 10% of the threshold). At the same time, calculate the difference between the actual load value of each exceeding component and the threshold, i.e., the exceeding limit amplitude, and arrange the exceeding limit amplitude data by component type to form a hardware load exceeding limit statistical report.

[0120] Step S1436: According to the number of exceeding components and the exceeding limit amplitude, refer to the correlation between load exceeding and fault occurrence in the hardware fault history data of the communication module to determine the hardware load exceeding limit risk level.

[0121] Retrieve the hardware fault history data of the communication module and analyze the probability of fault occurrence under different exceeding component quantities and different exceeding amplitudes. For example, when 1 component is slightly exceeding the limit, the probability of fault is low; when 2 or more components are seriously exceeding the limit, the probability of fault significantly increases.

[0122] According to the current number of out-of-limit components and the out-of-limit amplitude, a similar scenario in the fault history data is matched to determine a corresponding fault probability range, and a risk level is divided according to the fault probability range (such as low risk, medium risk, and high risk). The low risk corresponds to a fault probability lower than a preset low threshold, the medium risk corresponds to a fault probability between the low threshold and a high threshold, and the high risk corresponds to a fault probability higher than the high threshold.

[0123] Step S1437: integrating the hardware load out-of-limit risk level, the number of out-of-limit components, and the out-of-limit amplitude into a hardware load pre-judgment sub-result.

[0124] The hardware load out-of-limit risk level, the specific types and number of out-of-limit components, and the out-of-limit amplitude and corresponding load threshold of each out-of-limit component are integrated to form a hardware load pre-judgment sub-result. The high-risk out-of-limit component and its potential impact on hardware stability are explicitly indicated in the sub-result.

[0125] Step S144: extracting a core scheduling link in a software process scheduling sequence from the dynamic path, the core scheduling link being a link involving software dependent process startup or shutdown.

[0126] The software process scheduling sequence in the dynamic path is traversed to filter out links involving startup or shutdown operations of major software dependent processes and minor software dependent processes, which directly affect the core function implementation of the software system. For example, the startup link of the instruction parsing process and the shutdown link of the resource allocation process are both core scheduling links. The operation type (startup / shutdown), the process ID involved, the planned execution time, and the associated hardware adjustment node (if any) of the core scheduling link are recorded.

[0127] Step S145: analyzing the current software dependency relationship of the communication module, the current software dependency relationship of the communication module including inter-process data interaction dependency, resource sharing dependency, and execution order dependency.

[0128] All software processes currently running are traversed by a process dependency analysis tool to extract the dependency relationship between processes. The data interaction dependency refers to the transfer of data between processes through pipes, message queues, etc., such as the instruction parsing process sending parsed data to the data processing process. The resource sharing dependency refers to multiple processes sharing memory areas, file handles, etc., such as multiple processes accessing a configuration file simultaneously. The execution order dependency refers to a process that must be started after another process has completed execution, such as a data storage process that must be started after a data processing process has been started.

[0129] The above dependency relationships are presented in the form of a dependency graph, with nodes representing processes and edges representing dependency relationships. The attributes of the edges are annotated with dependency types to form a current software dependency graph.

[0130] Step S146: Simulate the software process scheduling process corresponding to the core scheduling link, analyze the integrity of the scheduling software dependency relationship, and determine whether there is a process dependency fracture or resource competition conflict.

[0131] Step S1461: Present the current software dependency relationship of the communication module in the form of a dependency graph, wherein the nodes in the dependency graph represent software processes, and the edges represent the dependency relationship between processes, and the attributes of the edges mark the dependency type and dependency strength.

[0132] Based on the current software dependency graph, a dependency graph is constructed, each node contains process ID, process name and current state attribute, and each edge contains dependency type (data interaction, resource sharing, execution sequence) and dependency strength attribute. The dependency strength is determined according to the interaction frequency between processes, the data size and the necessity of dependency. The more frequent the interaction, the larger the data size, the more necessary the dependency, the higher the strength.

[0133] Step S1462: Locate the software dependency process nodes involved in the core scheduling link in the dependency graph, and mark the software dependency process nodes to be closed and the software dependency process nodes to be started.

[0134] According to the process ID in the core scheduling link, the corresponding software dependency process node is found in the dependency graph, and the nodes to be closed and the nodes to be started are distinguished by different colors or marks. For example, the nodes to be closed are marked with red, and the nodes to be started are marked with green, and the planned operation time and operation type are marked beside the nodes.

[0135] Step S1463: Simulate closing the software dependency process node to be closed, delete all dependency edges corresponding to the software dependency process node, and analyze whether there is an isolated node in the remaining dependency graph. The isolated node is a software dependency process node that loses all input dependency edges or output dependency edges. The existence of an isolated node indicates that there may be a process dependency fracture.

[0136] The simulation closing operation is performed in the dependency graph, and all input and output dependency edges of the node to be closed are deleted. Traverse the remaining nodes, check whether each node has an input dependency edge (edge from other nodes to the node) or an output dependency edge (edge from the node to other nodes), if a node has neither an input dependency edge nor an output dependency edge, it is determined to be an isolated node. Record the process ID, process name and original dependency relationship of the isolated node to determine the location of the dependency fracture.

[0137] Step S1464: Simulate starting the software dependent process node to be started, add all dependent edges corresponding to the software dependent process node, analyze whether there is a same resource node being simultaneously pointed to by multiple software dependent process nodes in the new dependent graph, the resource node being an inter-process shared memory area, a file handle or a port resource, and multiple software dependent process nodes pointing to the same resource node indicating that there may be a resource contention conflict.

[0138] The simulation starting operation is performed in the dependent graph, the node to be started and all input and output dependent edges corresponding thereto are added, and a new dependent graph is formed. Resource nodes (such as shared memory nodes and file handle nodes) in the new dependent graph are identified, and it is checked whether two or more software dependent process nodes simultaneously point to the same resource node and the operation types (such as reading and writing) of the process nodes can cause a conflict (such as simultaneous writing operation). If such a situation exists, the identification and type (such as shared memory and file handle) of the resource node, the software dependent process node ID and the operation type involved are recorded to clearly indicate the specific location and conflict type of the resource contention conflict.

[0139] Step S1465: Dependence tracing is performed on the existing isolated node to determine the specific link of the dependence breakage, and the type of the broken dependence and the software dependent process node involved are recorded.

[0140] For the identified isolated node, the dependence graph structure before the closing operation is performed is traced back, and the source node and the target node corresponding to the original input dependent edge and the output dependent edge are found. By comparing the changes in the dependent graph before and after the closing operation, the link of the dependence breakage caused by the closing of the node to be closed is determined, for example, the isolated node originally depends on the data provided by the node to be closed, and after the node to be closed is deleted, the data dependence is broken.

[0141] The type of the broken dependence, such as data interaction dependence breakage and execution sequence dependence breakage, is recorded, and the ID and name of the isolated node and the original dependent node to be closed are recorded to form a dependence breakage detail record.

[0142] Step S1466: Resource occupation analysis is performed on the existing resource contention conflict to determine the type of the conflict resource and the competing process node, and the probability of the conflict occurrence is recorded.

[0143] For the resource node with the resource contention conflict, the occupation demand of each competing software dependent process node for the resource is analyzed, including the occupation time length, the access frequency and the operation permission (read / write). According to the priority of the process, the resource occupation demand and the processing record of the historical resource contention event, the probability of the current conflict occurrence is evaluated, for example, when a high-priority process and a low-priority process simultaneously request to write the same file handle, the conflict probability is high.

[0144] The specific type of the conflict resource, the detailed information of the competing process node, and the conflict probability obtained by the evaluation are recorded.

[0145] Step S1467: The dependency breakage analysis result, the resource competition conflict analysis result, and the corresponding record information are integrated into a software dependency pre-judgment sub-result.

[0146] The dependency breakage detailed record and the resource competition conflict analysis record are associated and integrated to determine the influence degree of the dependency breakage and the resource competition conflict on the software dependency relationship integrity. If there are multiple dependency breakages or high-probability resource competition conflicts, the software dependency relationship integrity is poor; if there are only a small amount of slight dependency breakages and no resource competition conflicts, the software dependency relationship integrity is good.

[0147] The finally formed software dependency pre-judgment sub-result includes the number of dependency breakages, the breakage link details, the number of resource competition conflicts, the conflict resource information, the conflict probability, and the software dependency relationship integrity evaluation conclusion.

[0148] Step S147: Extracting a key adaptation item in the external interaction adaptation strategy from the dynamic path, the key adaptation item being an item involving external link node protocol adjustment or data transmission mechanism change.

[0149] Traverse the external interaction adaptation strategy in the dynamic path, and filter out the adaptation items involving external link node communication protocol version change, protocol parameter adjustment (such as transmission window size, retransmission mechanism), and data transmission mechanism change (such as cache mechanism enabling / disabling, data fragmentation strategy adjustment, transmission priority change, etc.). Mark the above adaptation items as key adaptation items, and record their corresponding external link node ID, adaptation content, and planned execution time.

[0150] Step S148: Simulating the external interaction process corresponding to the key adaptation item in combination with the current external interaction link state of the communication module, and analyzing the interaction protocol compatibility and data transmission stability.

[0151] First, obtain the protocol version, negotiation parameter, data transmission rate, and error rate of the communication module, and other state parameters of the current external interaction link, which are used as the initial state of the simulation interaction process. For each key adaptation item, reproduce the external interaction scene in the simulation environment, for example, for the protocol adjustment type adaptation item, simulate the protocol negotiation process of the devices at both ends of the link; for the data transmission mechanism change type adaptation item, simulate the data sending, receiving, and cache processing process.

[0152] When analyzing the compatibility of the interaction protocol, check whether there are handshake failures, parameter mismatches, etc. in the protocol negotiation process, record the protocol compatibility status (fully compatible, partially compatible, incompatible) and the specific parameters of the incompatible. When analyzing the stability of data transmission, statistics the data loss rate, retransmission times and transmission delay changes in the simulation process, compare the transmission stability index differences before and after adaptation, and determine the influence of the key adaptation items on the transmission stability.

[0153] Step S149: Quantify the hardware load overrun risk, software dependency integrity judgment result and external interaction stability analysis result into comparable risk levels respectively, integrate the risk levels and refer to the stability records of the matching dynamic path in the historical state switching database to determine the path adaptation degree of the current dynamic path.

[0154] Set quantitative standards for hardware load overrun risk, software dependency integrity risk and external interaction stability risk, for example, divide the hardware load overrun risk into three levels of low, medium and high, which correspond to different overrun component quantities and overrun amplitudes; divide the software dependency integrity risk into four levels according to the number of dependency breaks and resource competition conflict probability; divide the external interaction stability risk into three levels according to the protocol compatibility level and data transmission index difference.

[0155] According to the analysis results, assign corresponding risk level quantitative values to each risk item, and weight sum the three quantitative values according to the preset weight to get the comprehensive risk value. Retrieve the historical path records in the historical state switching database that match the event type, impact level and adjustment direction of the current dynamic path, extract their corresponding relationship between the comprehensive risk value and the actual switching effect, and determine the path adaptation degree of the current dynamic path by comparing the current comprehensive risk value with the historical records. The adaptation degree ranges from 0 to 1, and the higher the value, the more suitable the path is to the current scenario.

[0156] Step S1410: Based on the path adaptation degree and the stable period after switching of the matching path in the history, estimate the stable period after execution of the current dynamic path, and generate a prediction result containing the path adaptation degree and the estimated stable period after switching.

[0157] Statistical the stable period after switching of the matching path in the history with similar path adaptation degree to the current dynamic path, calculate the average value of these periods as the reference stable period. According to the difference between the current path adaptation degree and the reference adaptation degree, correct the reference stable period, for example, if the current adaptation degree is higher than the reference adaptation degree, appropriately lengthen the estimated stable period; otherwise, shorten it.

[0158] The finally generated prediction result contains the path adaptation degree value, the estimated length of the stable period after switching, the risk level of each risk item and the specific analysis basis.

[0159] Step S150: According to the pre-judgment result, the dynamic path is converted into a state switching execution instruction, and a switching process data is generated by inputting the communication module to execute the switching operation.

[0160] According to the path adaptation degree in the pre-judgment result, it is judged whether to execute the dynamic path. If the adaptation degree meets the requirements, it is directly converted into an execution instruction; if it does not meet the requirements, the dynamic path is first modified and then the instruction is converted, and the switching process data is formed by real-time data collection during the execution process.

[0161] Step S151: Analyze the path adaptation degree in the pre-judgment result. If the path adaptation degree meets the preset adaptation standard, a state switching execution instruction is directly generated based on the dynamic path. If the path adaptation degree does not meet the preset adaptation standard, the hardware parameter adjustment sequence, software process scheduling order and external interaction adaptation strategy in the dynamic path are modified according to the hardware load overrun risk, software dependency fracture risk and external interaction instability risk in the pre-judgment result.

[0162] The preset adaptation standard is that the path adaptation degree is greater than or equal to the preset threshold. If the current path adaptation degree meets the standard, it directly enters the instruction generation link. If it does not meet the standard, for the hardware load overrun risk, the adjustment amplitude or order in the hardware parameter adjustment sequence is adjusted to reduce the load of the overrun component; for the software dependency fracture risk, the software process scheduling order is optimized, and the process start / closure logic is adjusted to avoid dependency fracture; for the external interaction instability risk, the protocol parameters or transmission mechanism in the external interaction adaptation strategy are modified to improve compatibility and stability.

[0163] During the modification process, the synergy of each adjustment item needs to be re-evaluated to ensure that each link of the modified dynamic path still maintains logical association and avoids the generation of new risks due to single risk modification.

[0164] Step S152: The modified dynamic path is structured according to the communication module instruction format requirements, and is converted into a state switching execution instruction containing hardware adjustment sub-instructions, software scheduling sub-instructions and external adaptation sub-instructions. Each sub-instruction contains an execution time identifier and a cooperative execution identifier.

[0165] The instruction format of the communication module follows the instruction set specification of its main control unit, and usually contains instruction header, instruction type, instruction parameter, check bit and instruction tail fields. During the structured processing, the hardware parameter adjustment sequence in the dynamic path is converted into hardware adjustment sub-instructions, each sub-instruction containing parameter identifier, adjustment value, adjustment duration, etc.; the software process scheduling order is converted into software scheduling sub-instructions, containing process ID, operation type, resource allocation parameters, etc.; the external interaction adaptation strategy is converted into external adaptation sub-instructions, containing link identifier, protocol parameter, cache configuration, etc.

[0166] Add execution timing identifier to each sub-instruction to clarify its execution time point in the overall switching process; add cooperative execution identifier to mark which sub-instructions need to be executed cooperatively, for example, the corresponding software scheduling sub-instruction can be triggered only after the execution of the hardware adjustment sub-instruction, to ensure that the switching operation is orderly and cooperatively performed.

[0167] Step S153: The state switching execution instruction is transmitted to the main control unit of the communication module, so that the main control unit of the communication module sequentially issues the hardware adjustment sub-instruction to the hardware control module, the software scheduling sub-instruction to the software management module, and the external adaptation sub-instruction to the external interaction module according to the execution timing identifier and the cooperative execution identifier.

[0168] The state switching execution instruction is transmitted to the main control unit through the internal bus. The main control unit first analyzes the execution timing identifier and the cooperative execution identifier in the instruction to construct an instruction execution timing table. According to the timing table, the main control unit first issues the hardware adjustment sub-instruction to the hardware control module. After receiving the instruction, the hardware control module drives the corresponding hardware components to complete parameter adjustment and returns execution feedback to the main control unit.

[0169] After receiving the feedback of the completion of the hardware adjustment, the main control unit issues the software scheduling sub-instruction to the software management module. The software management module returns execution feedback according to the start, close and resource allocation operations of the instruction execution process. Finally, the main control unit issues the external adaptation sub-instruction to the external interaction module according to the feedback of the completion of the software scheduling. The external interaction module performs protocol adjustment and data transmission mechanism configuration to complete the entire switching instruction issuing process.

[0170] Step S154: During the execution of the sub-instruction by each module, real-time collection of hardware adjustment parameter change data, software process state change data and external interaction link data is performed. After integration, switching process data is generated, which includes the execution time length of each sub-instruction, parameter comparison before and after execution and execution exception record.

[0171] When the hardware control module executes the sub-instruction, the dynamic change of the hardware parameter is collected in real time through ADC sampling, sensor monitoring and other methods. The adjustment start time, parameter value during adjustment and adjustment completion time are recorded. The execution time length is calculated, and the parameter difference before and after execution is compared. When the software management module executes the sub-instruction, the change of process state (such as from ready to running) and the change of resource occupancy rate are collected. The execution time length and execution result of process operation are recorded.

[0172] The external interaction module collects link data such as link protocol negotiation state, data transmission rate, and error code rate when executing the sub-instructions, records the execution duration and effect of the adaptation operation. If an exception occurs during the execution of a sub-instruction, such as hardware parameter adjustment failure, process startup timeout, etc., the occurrence time, type and code of the exception are recorded in real time. All collected data are classified and integrated according to the sub-instructions to form complete switching process data.

[0173] Step S160: Based on the stable period estimation in the switching process data and the pre-judgment result, the historical state switching database and the target associated network are updated to form a state switching closed-loop optimization identifier.

[0174] By comparing the switching process data with the pre-judgment result, the deviation between the switching effect and the estimation is analyzed, the related data is entered into the historical database, the node relationship of the target associated network is adjusted, and finally a closed-loop optimization identifier is generated to realize continuous optimization.

[0175] Step S161: Extracting the execution duration of each sub-instruction, the parameter comparison before and after execution, and the execution exception record from the switching process data, analyzing the actual switching effect.

[0176] For example, step S1611: separating the execution data of the hardware adjustment sub-instruction from the switching process data, extracting the execution duration of the hardware adjustment sub-instruction, comparing the change value of the hardware parameter before and after execution with the target change value of the hardware parameter adjustment sequence in the dynamic path, and calculating the hardware parameter adjustment deviation.

[0177] Separating the execution data of the hardware adjustment sub-instruction, classifying and arranging according to the parameter type, extracting the execution start and end time of each sub-instruction, and calculating the execution duration. Compare the actual change value of the hardware parameter after execution with the target change value set in the dynamic path, and calculate the difference value as the hardware parameter adjustment deviation. The smaller the deviation, the more in line with the expected hardware adjustment effect.

[0178] Step S1612: Statistics the execution exception record in the hardware adjustment process, determines the hardware component, type and duration of the exception, and analyzes the influence of the exception on the stable operation of the hardware.

[0179] Traverse the execution exception record of the hardware adjustment sub-instruction, count the number of exceptions according to the hardware component type, and determine the type of each exception, such as parameter adjustment overshoot, hardware response timeout, etc. Record the start and end time of each exception, and calculate the duration of the exception. Analyze the influence of the exception on the stable operation of the hardware, for example, voltage adjustment overshoot may cause temporary overload of the hardware component, affecting short-term stability; hardware response timeout may indicate that the component has a potential fault, affecting long-term stability.

[0180] Step S1613: Separate the execution data of the software scheduling sub-instruction from the switching process data, extract the execution time of the software scheduling sub-instruction, compare the active number, resource occupation rate of the software process before and after execution with the target value of the software process scheduling sequence in the dynamic path, and calculate the software process scheduling deviation.

[0181] Separate the execution data of the software scheduling sub-instruction, classify it according to the process operation type (start / close), extract the execution time of each sub-instruction. Compare the actual active number, CPU and memory resource occupation rate of the software process after execution with the target value in the dynamic path, calculate the difference value as the software process scheduling deviation, and evaluate the accuracy of the software scheduling operation.

[0182] Step S1614: Count the execution exception records in the software scheduling process, determine the software process where the exception occurs, the exception type and the exception recovery time, and analyze the influence of the exception on the software response speed.

[0183] Count the exception records in the software scheduling process, determine the software process ID and name involved in the exception, classify the exception type, such as process start failure, resource allocation shortage, etc. Record the time length from the occurrence to the recovery of the exception, analyze the influence of the exception on the software response speed, for example, process start failure will cause delay in execution of instructions dependent on the process, and resource allocation shortage will cause the software to respond slowly as a whole.

[0184] Step S1615: Separate the execution data of the external adaptation sub-instruction from the switching process data, extract the execution time of the external adaptation sub-instruction, compare the connection success rate, data transmission rate of the external interaction link before and after execution with the target value of the external interaction adaptation strategy in the dynamic path, and calculate the external interaction adaptation deviation.

[0185] Separate the execution data of the external adaptation sub-instruction, classify it according to the adaptation type (protocol adjustment / transmission mechanism change), extract the execution time. Compare the actual connection success rate, data transmission rate of the external interaction link after execution with the target value in the dynamic path, calculate the difference value as the external interaction adaptation deviation, and measure the execution effect of the external adaptation strategy.

[0186] Step S1616: Count the execution exception records in the external adaptation process, determine the external link node where the exception occurs, the exception type and the exception repair time, and analyze the influence of the exception on the stability of external interaction.

[0187] Count the exception records in the external adaptation process, determine the external link node identification where the exception occurs, the exception type, such as protocol negotiation failure, data buffer overflow, etc. Record the exception repair time, analyze the influence of the exception on the stability of external interaction, for example, protocol negotiation failure will cause link connection interruption, data buffer overflow will cause data loss, which will reduce the stability of external interaction.

[0188] Step S1617: The hardware parameter adjustment deviation, software process scheduling deviation, and external interaction adaptation deviation are compared with preset deviation thresholds of each link respectively to obtain deviation levels of each link. In combination with abnormal influence analysis results of each link, it is comprehensively judged whether the actual switching effect reaches the preset effect standard to generate an actual switching effect evaluation result. The actual switching effect evaluation result includes an effect level, deviation levels of each link, and an abnormality summary.

[0189] Preset deviation thresholds of each link are compared with corresponding thresholds to divide deviation levels (such as no deviation, slight deviation, moderate deviation, and serious deviation). In combination with abnormal influence analysis results of each link, if each link is no deviation or slight deviation and there is no serious abnormality, the actual switching effect reaches the preset standard; if there is moderate or serious deviation or there is a serious abnormality, the standard is not reached.

[0190] According to the evaluation result, an effect level (excellent, good, qualified, and unqualified) is divided, and deviation levels of each link, abnormal types, and influence analysis are integrated to form an actual switching effect evaluation result.

[0191] Step S162: The actual switching effect is compared with the post-switching stable period estimation in the pre-judgment result to calculate an estimation relative deviation. The estimation relative deviation is a difference between the actual stable period and the estimated stable period divided by the estimated stable period.

[0192] The stable running duration of the communication module after state switching is continuously monitored to determine the actual stable period. The actual stable period is compared with the estimated stable period in the pre-judgment result to calculate a difference value. The difference value is divided by the estimated stable period to obtain the estimation relative deviation. If the deviation is a positive value, it indicates that the actual stable period is longer than the estimation; if it is a negative value, it indicates that the actual stable period is shorter than the estimation. The greater the absolute value of the deviation, the lower the estimation accuracy.

[0193] Step S163: The actual execution effect of the hardware parameter adjustment sequence, the actual execution effect of the software process scheduling sequence, and the actual execution effect of the external interaction adaptation strategy are extracted from the switching process data, and the corresponding event hierarchical results and the current running state of the communication module are associated.

[0194] The actual stable duration after the hardware parameter adjustment, parameter deviation, and the like are extracted from the switching process data as the actual execution effect of the hardware adjustment sequence. The actual active state of the software process, resource occupancy, and the like are extracted as the actual execution effect of the software scheduling sequence. The connection success rate of the external link, transmission rate, and the like are extracted as the actual execution effect of the external adaptation strategy.

[0195] The actual execution effect is associated and bound with the event layering result (event ID, influence level, response priority) triggering the current state switching and the current running state (hardware parameters, software process state, link state) of the communication module before switching, to ensure data traceability.

[0196] Step S164: The event layering result, the current running state of the communication module, the actual executed dynamic path, the actual switching effect and the estimated deviation are recorded in the historical state switching database, and the historical event record, the historical switching path record and the historical stability record after switching are updated.

[0197] The event layering result and the associated running state information of the current event are added in the historical event record table, the actual executed dynamic path details including the corrected hardware adjustment sequence, the software scheduling sequence and the external adaptation strategy are recorded in the historical switching path record table, and the actual switching effect evaluation result, the actual stable period and the estimated relative deviation are recorded in the historical stability record after switching.

[0198] During the updating process, the events ID are correctly associated among the tables, and the incremental updating method is adopted to avoid covering the original historical data, so as to ensure the integrity and continuity of the database.

[0199] Step S165: The nodes corresponding to the current event layering result and the current running state of the communication module are searched from the target associated network, and the associated edge weight between the nodes is adjusted based on the actual switching effect.

[0200] Through the node searching function of the target associated network, the corresponding event level node, state element node and associated historical data node are searched according to the event ID and running state element identification of the current event layering result. If the actual switching effect is excellent, it indicates that the association relationship between these nodes is relatively accurate, and the weight of the associated edge between the nodes is increased; if the actual switching effect is poor, it indicates that there is deviation in the association relationship, and the weight of the associated edge is reduced, or the association relationship between the nodes is adjusted according to the actual executed dynamic path.

[0201] Step S166: If the actual switching effect is better than the historical similar record, the dynamic path is marked as the preferred path in the target associated network; if the actual switching effect is abnormal, the corresponding target associated node is marked in the target associated network.

[0202] Retrieving the historical records in the history state switching database that are the same as the current event type and impact level, comparing the actual switching effect with the effect level of the historical records of the same type, and if the actual switching effect level is higher than the highest level of the historical records of the same type, adding a “preferred path” mark to the corresponding dynamic path node in the target associated network, and recording the key parameters of the excellent effect; if the actual switching effect has serious abnormalities, such as hardware adjustment failure, frequent link interruption, etc., adding an “abnormal association” mark to the corresponding event level node and state element node, reminding the subsequent path generation to focus on the accuracy of the association relationship of these nodes. At the same time, record the reasons for the abnormal association, such as hardware sensitive parameter identification deviation, unreasonable software dependent process priority setting, etc.

[0203] Step S167: integrating the update records of the history state switching database and the adjustment records of the target associated network, generating a state switching closed-loop optimization report, and forming a state switching closed-loop optimization mark based on the state switching closed-loop optimization report, the state switching closed-loop optimization mark including optimization time, optimization content and expected effect after optimization.

[0204] Collecting the incremental update records of the history state switching database, including new event records, switching path records and stability records, and at the same time, collecting the node adjustment, associated edge weight modification and mark addition records of the target associated network. Classify and organize the above records in chronological order and by optimization type to form a state switching closed-loop optimization report.

[0205] The state switching closed-loop optimization report needs to specify the event type involved in this optimization, the details of the adjusted dynamic path, the specific changes of the target associated network, and the potential impact of these changes on subsequent state switching. Based on the report content, a state switching closed-loop optimization mark is generated, the optimization time is recorded as the timestamp of the report generation; the optimization content is clearly outlined as a brief summary of the database update items and network adjustment items; the expected effect after optimization is described as improving the adaptability of the dynamic path under the same event, shortening the state switching response time, reducing the probability of abnormality after switching, etc. The state switching closed-loop optimization mark will be used as the basis for calling the optimized data and network for the next state switching.

[0206] Based on the same inventive concept, please refer to Figure 2 , shows the structure schematic block diagram of the communication module state switching system 100 based on event triggering for executing the above-mentioned patrol video stream processing method provided by the embodiments of the application, which can include a communication unit 110, a machine readable storage medium 120 and a processor 130.

[0207] In this embodiment, the machine readable storage medium 120 and the processor 130 are both located in the event-triggered based communication module state switching system 100 and are separately arranged. However, it should be understood that the machine readable storage medium 120 can also be independent of the event-triggered based communication module state switching system 100 and can be accessed by the processor 130 through a bus interface. Alternatively, the machine readable storage medium 120 can also be integrated into the processor 130 and can communicate with external systems through the communication unit 110.

[0208] The processor 130 is the control center of the event-triggered based communication module state switching system 100, connects various parts of the event-triggered based communication module state switching system 100 through various interfaces and lines, executes various functions of the event-triggered based communication module state switching system 100 and processes data by running or executing software programs and / or modules stored in the machine readable storage medium 120 and calling data stored in the machine readable storage medium 120, thereby overall monitoring the event-triggered based communication module state switching system 100. Optionally, the processor 130 can include one or more processing cores; for example, the processor 130 can integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface and an application program, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor. Among them, the machine readable storage medium 120 is used to store machine executable instructions for executing the scheme of the present application, and the processor 130 is used to execute the machine executable instructions stored in the machine readable storage medium 120 to realize the video stream processing method provided by the foregoing method embodiment.

[0209] It should be noted that, in order to simplify the expression of the present disclosure and help understand one or more embodiments of the present application, in the foregoing description of the embodiments of the present application, various features are sometimes combined into one embodiment, figure or description thereof.

Claims

1. An event-triggered communication module state switching method, characterized in that, The method comprises: Collecting multi-source event source information in the running process of the communication module, performing event influence level division processing on the multi-source event source information to obtain event layering results, and the multi-source event source information includes hardware running abnormal signal, software instruction interaction information and external device association request information; Obtaining the current running state and the historical state switching database of the communication module, combining the event layering results and the current running state of the communication module to generate a target association network, and the historical state switching database stores historical event types, corresponding switching paths and stability records after switching; Based on the target association network and the stability records in the historical state switching database, a dynamic path for state switching of the communication module is generated, and the dynamic path includes a hardware parameter adjustment sequence, a software process scheduling sequence and an external interaction adaptation strategy; Performing state switching stability prediction processing on the dynamic path, combining the current hardware load threshold and the software dependency relationship of the communication module to generate a prediction result; According to the prediction result, the dynamic path is converted into a state switching execution instruction, and the switching operation is input to the communication module to generate switching process data; Based on the switching process data and the stability cycle estimation in the prediction result, the historical state switching database and the target association network are updated to form a state switching closed-loop optimization identifier.

2. The method of claim 1, wherein, The method comprises: Collecting hardware running abnormal signal from the hardware monitoring port of the communication module, and the hardware running abnormal signal includes power module voltage fluctuation, transmission module signal attenuation and interface module connection frequency change; Extracting software instruction interaction information from the software log of the communication module, and the software instruction interaction information includes instruction sending interval, instruction execution delay and instruction retry number; Collecting external device association request information from the external interaction port of the communication module, and the external device association request information includes request initiating device type, request data volume and request response time limit requirement; Performing influence range analysis on the hardware running abnormal signal to determine the number of hardware components affected by the signal and the importance of the components, and generating a hardware influence coefficient; Performing execution influence analysis on the software instruction interaction information to determine the interruption range of instruction abnormalities to the software process chain and the process recovery difficulty, and generating a software influence coefficient; Performing association influence analysis on the external device association request information to determine the influence breadth of request interruption to the external device interaction link and the link reconstruction cost, and generating an external influence coefficient; Integrating the hardware influence coefficient, the software influence coefficient and the external influence coefficient, referring to the preset influence level division standard of the communication module, generating an event layering result, and the event layering result includes an event core influence level, an association influence level and an indirect influence level, and different influence levels correspond to different state switching response priorities.

3. The method of claim 1, wherein, The acquisition communication module current running state and history state switching database, combined with the event hierarchical result and the communication module current running state, generate the target association network, including: Acquire the communication module current running state, the communication module current running state contains hardware component running parameter, software process active state and external interaction link state; Call the history state switching database, the history state switching database contains history event record, history switching path record and history switching after stability record, the history event record has corresponding relationship with the event type in the event hierarchical result, the history switching path record contains history hardware parameter adjustment sequence, history software process scheduling order and history external interaction adaptation strategy; The event core influence level in the event hierarchical result and the hardware component running parameter in the communication module current running state are associated, the hardware sensitive parameter corresponding to the core influence level is identified, the hardware sensitive parameter is the most significant parameter in the hardware component running for event response; The association influence level in the event hierarchical result and the software process active state in the communication module current running state are associated, the software dependent process corresponding to the association influence level is identified, the software dependent process is the process that needs to be scheduled preferentially in the software system affected by the event; The indirect influence level in the event hierarchical result and the external interaction link state in the communication module current running state are associated, the external link node corresponding to the indirect influence level is identified, the external link node is the node that needs to adjust the adaptation strategy in the external interaction affected by the event; The hardware sensitive parameter, software dependent process, external link node and history event record, history switching path record in the history state switching database are mapped and associated, the corresponding relationship between event hierarchy, current state element and history switching data is presented in the form of graph, the target association network is generated, the target association network contains nodes and association edges, the node represents event level, state element and history switching path, the association edge represents the influence relationship and mapping relationship between nodes.

4. The method of claim 1, wherein, The target association network and the stability record in the history state switching database are used to generate the dynamic path of communication module state switching, including: Extract the hardware sensitive parameter corresponding to the event core influence level from the target association network, combine the history hardware parameter adjustment sequence under the same hardware sensitive parameter in the history state switching database, analyze the first influence difference of different adjustment sequences on hardware stability, the first influence difference is reflected by the hardware stability time length in the history switching after stability record; Extract the software dependent process corresponding to the association influence level from the target association network, combine the history software process scheduling order under the same software dependent process in the history state switching database, analyze the second influence difference of different scheduling orders on software response speed, the second influence difference is reflected by the software instruction execution delay in the history switching after stability record; extracting external link nodes corresponding to the indirect influence hierarchy from the target associated network, combining historical external interaction adaptation strategies under the same external link nodes in the historical state switching database, analyzing third influence differences of different adaptation strategies on link stability, the third influence differences being reflected by link interruption frequencies in the historical stability records after switching; standardizing the first influence difference, the second influence difference and the third influence difference, converting them into unified relative influence scores, and selecting adapted hardware parameter adjustment directions, software process scheduling priorities and external interaction adaptation modes for the current event based on the relative influence scores; converting the hardware parameter adjustment directions into specific hardware parameter adjustment sequences, including the order and adjustment amplitude gradient of parameter adjustment; converting the software process scheduling priorities into specific software process scheduling sequences, including the logic of process closing and starting and the proportion of process resource allocation; converting the external interaction adaptation modes into specific external interaction adaptation strategies, including interaction protocol adjustment modes and data transmission buffer mechanisms; integrating the hardware parameter adjustment sequences, the software process scheduling sequences and the external interaction adaptation strategies to generate a dynamic path for communication module state switching, and setting associated identifiers for each link in the dynamic path to reflect the synergistic relationship of hardware, software and external interaction adjustment.

5. The method of claim 1, wherein, The state switching stability prediction processing of the dynamic path is combined with the current hardware load threshold and software dependency relationship of the communication module to generate a prediction result, including: extracting key adjustment nodes in the hardware parameter adjustment sequence from the dynamic path, the key adjustment nodes being nodes with an adjustment amplitude greater than a set amplitude or affecting multiple hardware components; obtaining the current hardware load threshold of the communication module, which includes the maximum bearing voltage, maximum transmission rate and maximum heat dissipation capacity of each hardware component; simulating the hardware parameter adjustment process corresponding to the key adjustment nodes, calculating the actual load values of each hardware component after adjustment, comparing the actual load values with the current hardware load threshold of the communication module, and analyzing the hardware load overrun risk; extracting core scheduling links in the software process scheduling sequence from the dynamic path, the core scheduling links being links involving the start or shutdown of software dependent processes; analyzing the current software dependency relationship of the communication module, which includes data interaction dependency, resource sharing dependency and execution order dependency between processes; simulating the software process scheduling process corresponding to the core scheduling links, analyzing the integrity of the software dependency relationship after scheduling, and determining whether there is a process dependency rupture or resource competition conflict; extracting key adaptation items in the external interaction adaptation strategy from the dynamic path, the key adaptation items being items involving external link node protocol adjustment or data transmission mechanism change; combining the current external interaction link state of the communication module, simulating the external interaction process corresponding to the key adaptation items, and analyzing the interaction protocol compatibility and data transmission stability; quantify the hardware load overrun risk, the software dependency integrity judgment result and the external interaction stability analysis result into comparable risk levels, integrate the risk levels and refer to the stability records of the matching dynamic path in the historical state switching database to determine the path adaptation degree of the current dynamic path; based on the path adaptation degree and the post-switching stability period of the historical matching path, estimate the stability period after execution of the current dynamic path, and generate a pre-judgment result including the path adaptation degree and the post-switching stability period estimation.

6. The method of claim 5, wherein, The simulation of the hardware parameter adjustment process corresponding to the key adjustment node calculates the actual load values of the adjusted hardware components, compares the actual load values with the current hardware load threshold of the communication module, and analyzes the hardware load overrun risk, including: establish a load calculation model for each hardware parameter adjustment item corresponding to the key adjustment node, the load calculation model is constructed based on the parameter-load correlation of the hardware components, and is used to reflect the corresponding relationship between the parameter adjustment amount and the load change amount; input the adjustment amplitude and adjustment time length in the hardware parameter adjustment item into the load calculation model, and output the real-time load change curve of each hardware component in the adjustment process; extract the peak load value in the real-time load change curve as the actual load value of the adjusted hardware components; classify the actual load values according to the hardware component types, and compare them with the corresponding current hardware load thresholds of the communication module one by one; statistically analyze the number of hardware components whose actual load values exceed the current hardware load threshold of the communication module and the overrun amplitude, which is the difference between the actual load value and the current hardware load threshold of the communication module; determine the hardware load overrun risk level according to the number of overrun components and the overrun amplitude, and refer to the correlation between load overrun and fault occurrence in the hardware fault history data of the communication module; integrate the hardware load overrun risk level, the number of overrun components and the overrun amplitude into a hardware load pre-judgment sub-result.

7. The method of claim 5, wherein, The simulation of the software process scheduling process corresponding to the core scheduling link analyzes the integrity of the software dependency relationship after scheduling, and judges whether there is process dependency fracture or resource competition conflict, including: present the current software dependency relationship of the communication module in the form of a dependency graph, where the nodes in the dependency graph represent software processes, and the edges represent the dependency relationship between processes, and the attributes of the edges mark the dependency type and dependency strength; locate the software dependency process nodes involved in the core scheduling link in the dependency graph, and mark the software dependency process nodes to be closed and the software dependency process nodes to be started; simulate closing the software dependency process nodes to be closed, delete all dependency edges corresponding to the software dependency process nodes, analyze whether there are isolated nodes in the remaining dependency graph, the isolated nodes are software dependency process nodes that lose all input dependency edges or output dependency edges, and the existence of isolated nodes indicates that there may be process dependency fracture; The software-dependent process node to be started is simulated to add all dependent edges corresponding to the software-dependent process node, and it is analyzed whether the same resource node is simultaneously pointed to by multiple software-dependent process nodes in the new dependent graph, the resource node being an inter-process shared memory area, a file handle, or a port resource, and multiple software-dependent process nodes pointing to the same resource node indicating that there may be a resource competition conflict; Dependence tracing is performed on the existing isolated nodes to determine the specific link of dependence rupture, and the type of ruptured dependence and the software-dependent process nodes involved are recorded; Resource occupation analysis is performed on the existing resource competition conflicts to determine the type of conflict resources and the competing process nodes, and the probability of conflict occurrence is recorded; The dependence rupture analysis results, resource competition conflict analysis results, and corresponding record information are integrated into software-dependent prediction sub-results.

8. The method of claim 1, wherein, According to the prediction results, the dynamic path is converted into state switching execution instructions, input into the communication module to perform switching operations, and switching process data is generated, including: Analyzing the path adaptation degree in the prediction results, if the path adaptation degree meets the preset adaptation standard, then directly generating state switching execution instructions based on the dynamic path; If the path adaptation degree does not meet the preset adaptation standard, then according to the hardware load overrun risk, software dependence rupture risk, and external interaction instability risk in the prediction results, modifying the hardware parameter adjustment sequence, software process scheduling order, and external interaction adaptation strategy in the dynamic path; The modified dynamic path is structured according to the communication module instruction format requirements, converted into state switching execution instructions containing hardware adjustment sub-instructions, software scheduling sub-instructions, and external adaptation sub-instructions, each sub-instruction containing an execution time identifier and a cooperative execution identifier; The state switching execution instructions are transmitted to the main control unit of the communication module, so that the main control unit of the communication module sequentially issues hardware adjustment sub-instructions to the hardware control module, software scheduling sub-instructions to the software management module, and external adaptation sub-instructions to the external interaction module according to the execution time identifier and the cooperative execution identifier; In the process of executing the sub-instructions in each module, real-time collection of hardware adjustment parameter change data, software process state change data, and external interaction link data is performed, and after integration, switching process data is generated, which contains the execution time of each sub-instruction, parameter comparison before and after execution, and execution exception record.

9. The method of claim 1, wherein, Based on the switching process data and the stable period estimation in the prediction results, the historical state switching database and the target associated network are updated to form a state switching closed-loop optimization identifier, including: Extracting the execution time of each sub-instruction, the parameter comparison before and after execution, and the execution exception record from the switching process data, and analyzing the actual switching effect; Comparing the actual switching effect with the stable period estimation after switching in the prediction results, calculating the estimation relative deviation, which is the difference between the actual stable period and the estimated stable period divided by the estimated stable period; extracting actual execution effects of the hardware parameter adjustment sequence, actual execution effects of the software process scheduling sequence, and actual execution effects of the external interaction adaptation strategy from the switching process data, correlating the corresponding event hierarchical results and the current running state of the communication module; recording the event hierarchical results, the current running state of the communication module, the actually executed dynamic path, the actual switching effect, and the estimated deviation into the historical state switching database, updating the historical event record, the historical switching path record, and the historical stability record after switching; finding nodes corresponding to the current event hierarchical results and the current running state of the communication module from the target associated network, and adjusting the associated edge weight between the nodes based on the actual switching effect; if the actual switching effect is better than the historical similar record, marking the dynamic path as the preferred path in the target associated network; if the actual switching effect is abnormal, marking the corresponding target associated node in the target associated network; integrating the update record of the historical state switching database and the adjustment record of the target associated network to generate a state switching closed-loop optimization report, forming a state switching closed-loop optimization mark based on the state switching closed-loop optimization report, the state switching closed-loop optimization mark including optimization time, optimization content, and expected effect after optimization.

10. An event-triggered based communication module state switching system, characterized in that, comprise: a processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the machine-executable instructions to perform the event-triggered communication module state switching method of any one of claims 1 to 9.

Images