Low-power control method and device for embedded system

Abstract

The application relates to the technical field of Internet of Things and Internet, and discloses a low-power consumption control method and device for an embedded system; the method is applied to a device comprising a main processing unit, an auxiliary processing unit, a low-power consumption monitoring module and a nonvolatile storage unit. The main processing unit and the auxiliary processing unit are connected through a communication interface and share the storage unit, and the auxiliary processing unit is connected with the low-power consumption monitoring module. The main processing unit generates running parameters with check information, the auxiliary processing unit checks and initializes the monitoring module after storage, then the main unit sends running state data, the auxiliary unit judges the sleep condition in combination with a wake-up threshold value and generates an instruction, the main unit responds to the instruction, stores key data in layers and feeds back, and the auxiliary unit checks and cuts off the power supply, and enters a low-power consumption standby mode. The scheme breaks through the traditional sleep mode, significantly reduces the standby power consumption, and meets the long-endurance requirement of Internet of Things devices.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things and Internet, in particular, to a low-power control method and device for embedded system. BACKGROUND

[0002] In the field of Internet of Things and Internet, embedded systems are widely used in portable battery-powered devices, wearable devices, and 4G communication devices, etc. Such devices have high demand for endurance. Current powerful embedded systems (hereinafter referred to as "large systems") have high difficulty in low-power design due to complex hardware modules, high energy consumption of computing and communication functions.

[0003] The prior art relies on the sleep mode of the large system to reduce power consumption, but due to the architecture limitation, the core module still needs to maintain basic power supply in the sleep state, which has high standby power consumption. The low-power effect cannot meet the long-endurance demand of the device, resulting in frequent replacement or charging of the battery, which seriously affects the user experience, especially in industrial field monitoring and other scenes that need long-term unattended, thereby restricting the stable application and promotion of the device. SUMMARY

[0004] The main purpose of the present application is to provide a low-power control method for embedded system, which aims to solve the technical problem that the powerful embedded "large system" in the field of Internet of Things and the like relies on its own sleep mode to reduce power consumption, which has poor effect, high standby power consumption, and cannot meet the long-endurance demand of the device.

[0005] The first aspect of the present application provides a low-power control method for embedded system, which is applied to a low-power control device for embedded system. The device includes a main processing unit, an auxiliary processing unit, and a low-power monitoring module. The main processing unit and the auxiliary processing unit are connected through a communication interface and are electrically connected to a non-volatile storage unit. The auxiliary processing unit is electrically connected to the low-power monitoring module. The method comprises:

[0006] The main processing unit generates running parameters containing check information, which are transmitted to the auxiliary processing unit through the communication interface. After the check is passed by the auxiliary processing unit, the running parameters are stored in the non-volatile storage unit. The auxiliary processing unit initializes the low-power monitoring module according to the running parameters and feeds back a completion signal to the main system, wherein the main system is the core computing system in the embedded system dominated by the main processing unit.

[0007] The main processing unit sends running state data containing task state, environmental data, and time limit information to the auxiliary processing unit. The auxiliary processing unit determines the sleep condition by combining the wake-up threshold in the stored running parameters and the running state data, and generates corresponding instructions to transmit to the main processing unit.

[0008] The main processing unit writes the key data before hibernation to the non-volatile storage unit in layers in response to the hibernation instruction, and feeds back the data identification and the log, which is verified by the auxiliary processing unit, and the auxiliary processing unit controls the power supply loop of the main processing unit to be cut off, and the auxiliary processing unit enters a low-power consumption state after binding the related data, wherein the key data is the data necessary for the system to recover to the state before hibernation; the related data is the data associated with the state parameters before hibernation of the main processing unit;

[0009] After the auxiliary processing unit monitors the wake-up event and verifies it, the auxiliary processing unit controls the power supply loop of the main processing unit to be turned on, transmits the wake-up information and the data storage location to the main processing unit, and the main processing unit reads the key data before hibernation to recover to the state before hibernation;

[0010] The main processing unit extracts features based on historical data of multiple hibernation-wake-up cycles, calculates an optimization coefficient to generate an update set, and synchronizes the update set to the auxiliary processing unit, and the auxiliary processing unit updates the stored operating parameters and feeds back to the main processing unit.

[0011] Further, the main processing unit sends operating state data containing task state, environmental data and time limit information to the auxiliary processing unit, and the auxiliary processing unit determines the hibernation condition by combining the wake-up threshold in the stored operating parameters and the operating state data, and generates a corresponding instruction to the main processing unit, including:

[0012] The main processing unit sends operating state data containing task state, environmental data, credibility identification, time limit information and task planning to the auxiliary processing unit;

[0013] The auxiliary processing unit verifies the credibility identification in the operating state data, and comprehensively determines whether the preset hibernation condition is met by combining the wake-up threshold, the effective time limit information and the task planning in the stored operating parameters;

[0014] The auxiliary processing unit generates an instruction containing data saving identification, credibility and time limit verification result and task planning association description according to the determination result;

[0015] The main processing unit receives the instruction, confirms the data validity and the task planning matching according to the verification result, executes the instruction and feeds back the result to the auxiliary processing unit.

[0016] Further, the step of the auxiliary processing unit generating an instruction containing data saving identification, credibility and time limit verification result and task planning association description according to the determination result, including:

[0017] According to the determination result, a data storage mark containing a storage path, a data type and a storage priority is generated, and the data is distinguished into mandatory data and optional data, wherein the mandatory data is the data necessary for the system to restore to the state before hibernation after hibernation, and the optional data is the supplementary data for auxiliary analysis or not necessary;

[0018] The credibility and timeliness verification result generates a verification digest, a status code and a report fragment;

[0019] The key nodes are extracted in combination with the task planning association description, and the association index and the time node are generated;

[0020] The data storage mark, the verification result and the key node description are integrated, and the integration verification information is added to form the instruction.

[0021] Further, the step of combining the task planning association description to extract the key nodes, generate the association index and mark the time nodes, comprises:

[0022] The task planning association description is parsed, and the task start, end, main task, subtask, resource allocation and state conversion node are identified as the key nodes;

[0023] The key nodes are assigned a unique identifier, a hierarchical association index is generated to mark the subordinate relationship, and a state association index is generated to record the conversion logic;

[0024] The execution time, planned execution point and state conversion expected time of each node are extracted as time nodes and marked;

[0025] The key nodes, hierarchical association index, state association index and time nodes are arranged into an association table according to the corresponding relationship;

[0026] The association table is arranged according to the subordinate relationship and the state conversion logic.

[0027] Further, the main processing unit responds to the hibernation instruction, and the key data before hibernation is written into the non-volatile storage unit in layers, and the data mark and the log are fed back, and after the auxiliary processing unit verifies, the auxiliary processing unit controls the power supply loop of the main processing unit to be cut off, and the auxiliary processing unit binds the related data and enters the low-power consumption state.

[0028] The main processing unit writes the key data before hibernation into the non-volatile storage unit in layers, generates the data mark and the storage log of each layer, and feeds back the data mark, the log digest and the layered storage result to the auxiliary processing unit;

[0029] The auxiliary processing unit verifies the data mark matching and the log digest integrity, and confirms that the data storage of each layer is completed, and then the auxiliary processing unit cuts off the power supply loop of the main processing unit;

[0030] The auxiliary processing unit binds the data identifier, the log digest and the main processing unit state parameter according to a data layer corresponding relationship and records the binding;

[0031] The auxiliary processing unit associates the data layer information with the binding record and carries the associated record into the low-power state.

[0032] Further, the step of the auxiliary processing unit binding the data identifier, the log digest and the main processing unit state parameter according to a data layer corresponding relationship and recording the binding comprises:

[0033] The data layer identifier is identified, the core layer and the extended layer are distinguished, the data identifier feature information and the log digest key information of each layer are extracted;

[0034] The core layer and the extended layer are respectively generated with a merged check value, the feature information, the key information, the check value and the corresponding layer main processing unit state parameter are bound, and a layer type label is added;

[0035] The layer identifier is added to each binding group as an association key, and the association record of the type label and the binding group is established;

[0036] The association record and all binding groups are stored according to the layer identifier sequence.

[0037] Further, the step of adding the layer identifier to each binding group as an association key and establishing the association record of the type label and the binding group comprises:

[0038] The unique layer identifier containing the layer information and the check bit is generated for each binding group as the association key embedded in the binding group;

[0039] The attribute features of the type label are identified and classified, and the features of the classified labels are extracted;

[0040] The type label and the corresponding binding group are associated by comparing the association key features and the label features, and the association record is generated;

[0041] The key check code is added to the association record, and the corresponding table of the label features and the association record is stored.

[0042] Further, the step of identifying the attribute features of the type label and classifying the classified labels and extracting the features of the classified labels comprises:

[0043] The type label is scanned, the basic attributes, dynamic and static attributes and association attribute features are extracted, and the association object and the association strength are determined;

[0044] The association-still dual-dimension classification matrix is formed according to the association strength classification and the division of dynamic and static attributes;

[0045] The common and differentiated features and the association feature coefficients are extracted for the same type labels in the classification matrix;

[0046] Correlate the hierarchical classification result with the feature extraction data, and bind the two-dimensional classification with the feature coefficient.

[0047] Further, the verification data identifies the matching and the log summary integrity, and after confirming that each layer of data storage is completed, the step of cutting off the power supply loop of the main processing unit by the auxiliary processing unit comprises:

[0048] Receiving the data identifier, the log summary and the hierarchical storage result, identifying each layer type identifier and the irreplaceable identifier, extracting the core layer containing the irreplaceable identifier and the corresponding type, and the rest as an expansion layer;

[0049] Extracting the storage time stamp and capacity information of the core layer and the expansion layer, and sorting them according to the time stamp and the capacity from large to small;

[0050] Assigning a verification sequence number containing a time stamp segment and a capacity ratio to each layer after sorting, and forming a hierarchical verification sequence by correlating the layer type identifier, the sorting result and the verification sequence number.

[0051] The second aspect of the application also proposes an embedded system low-power control device, comprising:

[0052] The embedded system low-power control device is applied to an embedded system low-power control device, which comprises a main processing unit, an auxiliary processing unit and a low-power monitoring module. The main processing unit and the auxiliary processing unit are connected through a communication interface, and both are electrically connected with a non-volatile storage unit. The auxiliary processing unit is electrically connected with the low-power monitoring module. The device comprises:

[0053] A parameter transmission and storage module, which comprises a parameter generation sub-module on the main processing unit side and a parameter verification and storage sub-module on the auxiliary processing unit side. The parameter generation sub-module is used to generate running parameters containing verification information and transmit them to the auxiliary processing unit through the communication interface. The parameter verification and storage sub-module is used to receive the running parameters and store them in the non-volatile storage unit after verification. The auxiliary processing unit initializes the low-power monitoring module according to the running parameters and feeds back a completion signal to the main system, wherein the main system is the core operation system in the embedded system dominated by the main processing unit.

[0054] A state monitoring and determination module, which comprises a state sending sub-module on the main processing unit side and a condition determination sub-module on the auxiliary processing unit side. The state sending sub-module is used to send running state data containing task state, environmental data and time limit information to the auxiliary processing unit. The condition determination sub-module is used to determine the hibernation condition by combining the wake-up threshold in the stored running parameters and the running state data, and generate corresponding instructions to transmit to the main processing unit.

[0055] The hibernation control module comprises a data storage submodule on the main processing unit side and a power supply control submodule on the auxiliary processing unit side; the data storage submodule is configured to write key data before hibernation into a non-volatile storage unit in layers and feed back data identification and a log in response to a hibernation instruction; the power supply control submodule is configured to check data identification matching and log integrity, control a power supply loop of the main processing unit to be cut off after verification, and control the auxiliary processing unit to enter a low-power-consumption state after relevant data is recorded, wherein the key data is data necessary for the system to be restored to a state before hibernation after hibernation; and the relevant data is data associated with a state parameter before hibernation of the main processing unit.

[0056] The wake-up recovery module comprises an event monitoring submodule on the auxiliary processing unit side and a state recovery submodule on the main processing unit side; the event monitoring submodule is configured to monitor a wake-up event and control a power supply loop of the main processing unit to be turned on and transmit wake-up information and a data storage location after verification; and the state recovery submodule is configured to read the key data before hibernation and restore to the state before hibernation.

[0057] The parameter update feedback module comprises a coefficient calculation submodule on the main processing unit side and a parameter update submodule on the auxiliary processing unit side; the coefficient calculation submodule is configured to extract features based on historical data of multiple hibernation-wake-up cycles, calculate optimization coefficients to generate an update set, and synchronize the update set to the auxiliary processing unit; and the parameter update submodule is configured to update stored operating parameters and feed back to the main processing unit.

[0058] The first aspect of the present application brings the following beneficial effects:

[0059] The present application realizes low-power-consumption control through cooperation of the main and auxiliary processing units: first, the main unit transmits operating parameters containing verification information, and the auxiliary unit stores the parameters and initializes the monitoring module; then, the auxiliary unit determines hibernation conditions based on the operating parameters and state data, generates an instruction, the main unit stores key data, and the auxiliary unit cuts off power supply of the main unit and waits in a low-power-consumption state; when woken up, the auxiliary unit turns on the power supply, and the main unit recovers the state, and can also optimize parameters based on historical data. The present application avoids the limitation of the sleep mode of a large system, greatly reduces standby power consumption, and effectively meets the long-endurance requirement of Internet of Things devices. BRIEF DESCRIPTION OF DRAWINGS

[0060] Figure 1 is a flowchart of a low-power-consumption control method of an embedded system according to an embodiment of the present application;

[0061] Figure 2 is a structural diagram of a low-power-consumption control device of an embedded system according to an embodiment of the present application;

[0062] Figure 3 is a structural diagram of a computer device according to an embodiment of the present application;

[0063] The objectives, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0064] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0065] Those skilled in the art can understand that the singular forms "a," "an," and "the" used herein include plural forms unless specifically stated otherwise. It should be further understood that the use of the term "include" in the specification of the present application means that a feature, integer, step, operation, element, module and / or assembly exists, but does not exclude the existence or addition of one or more other features, integers, steps, operations, elements, modules, assemblies and / or combinations thereof. It should be understood that when an element is "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or there can be an intermediate element. In addition, "connected" or "coupled" as used herein can include wireless connection or wireless coupling. The phrase "and / or" as used herein includes all or any combination of the associated listed items.

[0066] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have meanings consistent with those in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as such.

[0067] Reference Figure 1 The embodiment of the present application provides a low-power consumption control method for an embedded system, which is applied to a low-power consumption control device for an embedded system. The device comprises a main processing unit, an auxiliary processing unit and a low-power consumption monitoring module. The main processing unit and the auxiliary processing unit are connected through a communication interface and are electrically connected to a non-volatile storage unit. The auxiliary processing unit is electrically connected to the low-power consumption monitoring module. The method comprises the following steps:

[0068] S1: The main processing unit generates operation parameters containing check information, transmits them to the auxiliary processing unit through a communication interface, stores them in the non-volatile storage unit after passing the check of the auxiliary processing unit, and the auxiliary processing unit initializes the low-power monitoring module according to the operation parameters and feeds back a completion signal to the main system, wherein the main system is a core operation system in an embedded system dominated by the main processing unit;

[0069] S2: The main processing unit sends operation state data containing task state, environmental data and time limit information to the auxiliary processing unit, the auxiliary processing unit determines the sleep condition by combining the stored wake-up threshold in the operation parameters and the operation state data, generates corresponding instructions and transmits them to the main processing unit;

[0070] S3: The main processing unit responds to the sleep instruction, writes the key data before sleep into the non-volatile storage unit in layers, and feeds back data identification and logs, which are checked by the auxiliary processing unit, and then the auxiliary processing unit controls the power supply loop of the main processing unit to be cut off, and the auxiliary processing unit records the related data and enters the low-power state, wherein the key data is the data necessary for the system to recover to the state before sleep after sleep, including task execution progress, core configuration parameters, data storage address index, and is directly related to the operation state data before sleep; the related data is the associated data of data identification, log summary and state parameters of the main processing unit before sleep, and the state parameters of the main processing unit include power information, task completion mark and operation mode parameters;

[0071] S4: After the auxiliary processing unit monitors the wake-up event and verifies it, the auxiliary processing unit controls the power supply loop of the main processing unit to be turned on, transmits the wake-up information and data storage location to the main processing unit, and the main processing unit reads the key data before sleep to recover to the state before sleep;

[0072] S5: The main processing unit extracts features based on historical data of multiple sleep-wake cycles, calculates an optimization coefficient to generate an updated set and synchronizes it to the auxiliary processing unit, and the auxiliary processing unit updates the stored operation parameters and feeds back to the main processing unit.

[0073] In step S1, in the battery-powered industrial field 4G monitoring device, the embedded system low-power control method is applied to the embedded system low-power control device, which includes a main processing unit, an auxiliary processing unit, and a low-power monitoring module. The main processing unit and the auxiliary processing unit are connected through a communication interface and are electrically connected to a non-volatile storage unit. The auxiliary processing unit is electrically connected to the low-power monitoring module. The main system is the core operation system in the embedded system dominated by the main processing unit, forming a control closed loop. Step S1 is the basic preparation link of the low-power control of the dual system. After the device is powered on, the main processing unit (such as a high-performance MCU) that undertakes complex tasks such as video encoding and 4G communication first generates running parameters containing CRC check codes. The parameters include a 30-minute wake-up period, a temperature wake-up threshold (0-50°C), a data storage path, and other core configurations. They are packaged in the format of "parameter type-value-storage address" and are attached with EEPROM storage address identifiers (such as 0x0001-0x0008). They are transmitted to the auxiliary processing unit (low-power MCU) through the SPI interface. The auxiliary processing unit reorganizes the parameters after receiving them, verifies the CRC check codes, and stores them in the non-volatile storage unit (EEPROM, Electrically Erasable Programmable Read-Only Memory) according to the address identifiers after verification. Then, the parameters are read from the corresponding addresses in the EEPROM, repackaged in the original format, and a new CRC check code is calculated. The repackaged result is compared with the original transmission parameters and the two check codes. After confirming the consistency, the low-power monitoring module is initialized according to the parameters, i.e., setting a 30-minute timer (a timing wake-up trigger source) and a GPIO interrupt pin (connected to a temperature sensor for threshold-triggered wake-up). Finally, a "parameter initialization complete" signal is fed back to the main processing unit, providing accurate basic configurations for subsequent steps such as sleep judgment in S2 and sleep execution in S3. For example, the auxiliary processing unit in subsequent S2 determines whether to sleep based on the wake-up threshold stored in this step. This step ensures the accuracy of the transmission and storage of the running parameters through the verification of the check codes and the secondary comparison of the parameters, avoiding deviations in subsequent sleep judgment caused by parameter errors. The non-volatile storage ensures that the parameters are not lost after power failure, and the initialization of the low-power monitoring module lays the foundation for subsequent wake-up triggers, establishing a reliable starting link for the low-power control of the dual system and improving the stability of the system and the smoothness of the subsequent process.

[0074] In step S2, after completing the video acquisition and 4G uploading task, the main processing unit sends the running state data to the auxiliary processing unit: the task state containing "task completed", the environment data of the current environment temperature 25°C, the data reliability identifier (98%), the time limit information (the data is valid before the next task starts), and the task planning of "starting the next acquisition in 30 minutes". After receiving, the auxiliary processing unit first verifies the data reliability identifier to confirm that the data is reliable, then retrieves the running parameters (wake-up threshold 0-50°C, 30-minute wake-up period) stored in S1, and combines the running state data to determine that the current temperature 25°C is within the wake-up threshold range, the remaining valid duration of the data is 30 minutes, which matches the minimum sleep duration of 25 minutes, and the task planning shows that the interval between the next start time and the current time satisfies the sleep condition. Then generate a sleep instruction, clearly specify the storage requirements of "mandatory video index to / data / index (core layer data), optional storage of environment temperature log (extended layer data)", and attach the verification result and task planning association description, and transmit it to the main processing unit, providing a basis for the main processing unit to execute the sleep operation of S3. This step ensures that the sleep decision is scientific and accurate through multi-dimensional data verification and condition determination, avoids misjudgment leading to task interruption or power waste; the instruction clearly specifies the storage requirements and task association, providing clear guidance for subsequent data storage and wake-up recovery, and improving the system collaboration efficiency.

[0075] In step S3, the main processing unit receives the hibernation instruction sent by the auxiliary processing unit in S2, and immediately responds to execute: write the core layer data (video index) and the extension layer data (environment temperature log) to the 0x0100-0x0110 and 0x0200-0x0205 address segments of the EEPROM according to the instruction requirements, generate the layer data fingerprints (such as SHA-256 hash value) and storage logs (record storage time, address range), and feedback these information to the auxiliary processing unit. After receiving, the auxiliary processing unit first verifies the matching of the core layer data fingerprint with the preset algorithm result, then verifies the integrity of the extension layer log summary, and confirms that the two layer data are stored without error, and then triggers the MOS tube to cut off the power supply circuit of the main processing unit (the main unit is completely powered off, and the power consumption is reduced to 0). Subsequently, the auxiliary processing unit binds the data fingerprint, log summary and the task completion state parameter before the main unit hibernation, marks the "core layer" and "extension layer" type labels, carries the record into the low-power standby mode (the power consumption is only 50 μA), and reserves the key data association information for the wake-up recovery of S4, wherein the key data is the data necessary for the system to recover from the hibernation state to the state before hibernation, including task execution progress, core configuration parameter, data storage address index, and is directly related to the running state data before hibernation; the related data is the association data of data identification, log summary and the state parameter before the main processing unit hibernation, and the state parameter of the main processing unit includes power information, task completion mark, and running mode parameter. This step ensures the data integrity and reliability through hierarchical storage and double verification, completely cuts off the power supply of the main unit to maximize the power consumption, binds the record to provide the basis for subsequent wake-up recovery, and takes into account the low-power effect and system reliability.

[0076] In step S4, the auxiliary processing unit continuously monitors the 30-minute timer initialized in S1 and the temperature sensor in the low-power standby mode: when the timer triggers (reaches the 30-minute interval planned in S2), the auxiliary processing unit immediately records the triggering time (e.g., 10:30:00) and the triggering feature ("timed wake-up"). Then, a double verification is performed: the triggering time is compared with the "wake-up permission period" (00:00-24:00, the whole period is allowed) stored in S1 to confirm that it is within the permitted range; the triggering feature is matched with the "timed trigger" template in the "wake-up feature library" with a matching degree of 95%. After the verification passes, the auxiliary processing unit controls the MOS tube to turn on the main processing unit power supply circuit, and transmits the wake-up information (triggering time, feature matching result) and the data storage location recorded in S3 (core layer 0x0100-0x0110, extended layer 0x0200-0x0205). After the main processing unit is powered on, it reads the video index and state parameters according to the storage location, quickly recovers to the task completion state before S3 hibernation, and prepares for the next video acquisition. This step avoids false wake-up through double verification, ensures fast data reading through accurate transmission of storage location, realizes seamless system recovery, guarantees task continuity, and reduces power waste during the wake-up process.

[0077] In step S5, after the device has been running for 1 week, the main processing unit counts the historical data generated in the S3-S4 cycle: core data includes daily hibernation time (average 28 minutes / time), wake-up times (8 times / day during the day, 4 times / night); extended data includes environmental temperature fluctuations (25-35°C during the day, 15-20°C at night), task execution time (video upload average 45 seconds). The main processing unit extracts the fluctuation features: daytime wake-up interval fluctuation ±2 minutes, nighttime fluctuation ±5 minutes; marks abnormal points: a daytime wake-up time of 70 seconds (due to weak 4G signal). Compare these data with the S1 preset reference (standard hibernation time 30 minutes, wake-up time ≤50 seconds), calculate the combined optimization coefficient (daytime hibernation time adjusted to 25 minutes, nighttime to 35 minutes) and the abnormal correction coefficient (wake up 10 seconds earlier when the signal is weak). Add a time effectiveness label (validity period 7 days) and a "signal abnormality" type identifier to the coefficient, generate a hierarchical update set containing core / extended layer data, and synchronize it to the auxiliary processing unit through SPI. The auxiliary processing unit updates the running parameters in the EEPROM according to the identifier, and feeds back a "parameter update completion signal" to make the subsequent S2 hibernation determination more suitable for the actual scene. This step dynamically optimizes parameters through historical data analysis, adapts the low-power strategy to environmental and task changes, continuously improves the endurance, forms a "run- optimize-iterate" closed loop, and enhances the system's self-adaptation ability.

[0078] In one embodiment, the main processing unit generates the running parameters containing the check information, transmits them to the auxiliary processing unit through the communication interface, stores them in the non-volatile storage unit after the check by the auxiliary processing unit, and the auxiliary processing unit initializes the low-power monitoring module according to the running parameters and feeds back the completion signal to the main system, which comprises:

[0079] S10: generating the running parameters containing the check information, and transmitting them to the auxiliary processing unit through the communication interface after the main processing unit packs them according to the preset format and adds the storage address identifier;

[0080] S11: receiving and reorganizing the running parameters, verifying the check information by the auxiliary processing unit, and storing them in the non-volatile storage unit according to the address identifier after the check;

[0081] S12: reading the parameters from the corresponding address of the non-volatile storage unit, packing them according to the original format by the auxiliary processing unit, recalculating the check information, and comparing the packing result with the original transmission parameters and the check information twice;

[0082] S13: if the comparison is consistent, initializing the low-power monitoring module according to the running parameters and feeding back the completion signal to the main processing unit, otherwise requesting retransmission, which is operated by the auxiliary processing unit.

[0083] In this embodiment, first, the main processing unit generates the running parameters containing CRC32 check information, including 30-minute wake-up period, 0-50℃ temperature threshold, data storage path, etc., and packs them in the preset format of "parameter ID (1 byte) + value (4 bytes) + check bit (1 byte)", adds EEPROM storage address identifier to each parameter (such as 0x0001-0x0005 corresponding to the wake-up period), and transmits them to the auxiliary processing unit through the SPI interface. Next, the auxiliary processing unit reorganizes the parameter frame received into the original parameter structure, verifies the check information using the CRC32 algorithm, confirms that the check bit of the wake-up period parameter "0x1E (30 minutes)" matches, and then writes the parameter into the specified area of the EEPROM according to the address identifier. After that, the auxiliary processing unit reads the wake-up period parameter from the 0x0001 address, repacks it into "0x01 (ID) + 0x1E (value) + 0x7A (new check bit)" in the original format, compares the byte sequence of the newly packed result with the original transmission parameter, and simultaneously checks whether the original CRC32 value (0x1234) and the newly calculated value (0x1234) are consistent. Finally, since the comparison is consistent, the auxiliary processing unit initializes the low-power timer to a 30-minute period according to the parameter, configures the GPIO interrupt pin as the temperature sensor input mode, and sends a "0x00 (complete)" signal to the main processing unit; if the check is inconsistent (such as a certain value being transmitted as 0x1F by mistake), it returns a "0x01 (retransmit)" instruction to ensure that the subsequent S2 sleep determination can be based on accurate parameters. Through layered verification and bidirectional comparison, the accuracy of the running parameter transmission and storage is ensured, and deviations in subsequent processes caused by parameter errors are avoided; the address identifier and the retransmission mechanism ensure data integrity, lay a reliable foundation for collaborative control of the dual systems, and improve system stability.

[0084] In one embodiment, the steps of reading the parameter from the corresponding address of the non-volatile storage unit, packing it in the original format by the auxiliary processing unit and recalculating the check information, and comparing the packed result with the original transmission parameter and the check information twice, include:

[0085] S120: reading the parameter from the corresponding address of the non-volatile storage unit, packing it in the original format and recalculating the check information, and extracting the features of the packed parameter to generate a parameter feature value;

[0086] S121: encrypting the packed parameter, the recalculated check information, and the parameter feature value, and dividing the packed parameter into main parameters and auxiliary parameters and associating them with the check information;

[0087] S122: storing the encrypted content and preferentially verifying the matching of the main parameters and the check information;

[0088] S123: after the main parameter verification is passed, decrypting the auxiliary parameter to perfect the main parameter, and confirming the parameter consistency in combination with the parameter feature value.

[0089] In the industrial field 4G monitoring device, the sub-step is a key reinforcement link of parameter storage verification in this embodiment. The auxiliary processing unit reads the stored operating parameters (including 30-minute wake-up period, 0-50°C temperature threshold) from the 0x0001-0x0008 address of the EEPROM, repackages them in the original format "parameter ID + value + check bit", calculates new CRC32 check information (such as the new check value of the wake-up period parameter is 0x7A), and generates parameter characteristic values (such as the combined characteristic code 0xABC123 of "wake-up period + temperature threshold" obtained by the hash algorithm) by extracting features from the packaged parameters. Then, the auxiliary processing unit uses the AES algorithm to encrypt the packaged parameters, new check information, and characteristic values, divides the parameters into main parameters (wake-up period, temperature threshold core control items) and auxiliary parameters (storage path and other supplementary configurations), and associates the main parameters with the check information in priority (such as the wake-up period 0x1E binds the check value 0x7A). After storing the encrypted content, the auxiliary processing unit verifies the matching of the main parameters and the check information in priority, confirms that the wake-up period parameter value corresponds to the check value without error. After the main parameter verification passes, the auxiliary parameters (such as the / data / index storage path) are decrypted and called to complete the main parameter configuration, and finally the parameter characteristic value 0xABC123 is compared with the characteristic value generated during the original transmission to confirm the consistency of all parameters, ensuring that the parameters transmitted in S1 have not been tampered with, providing an absolutely reliable basis for subsequent S2 to determine sleep based on these parameters. Through encryption processing and hierarchical verification, the parameter security and integrity are improved, the main parameter verification in priority ensures the reliability of the core logic, and the characteristic value comparison further ensures the parameter consistency, which builds a strong defense line for stable system operation.

[0090] In one embodiment, the main processing unit sends running state data containing task status, environmental data, and time limit information to the auxiliary processing unit, and the auxiliary processing unit determines the sleep condition by combining the wake-up threshold in the stored operating parameters and the running state data, generates corresponding instructions and transmits them to the main processing unit, including:

[0091] S20: The main processing unit sends running state data containing task status, environmental data, credibility identifier, time limit information, and task planning to the auxiliary processing unit;

[0092] S21: The auxiliary processing unit verifies the credibility identifier in the running state data, and comprehensively determines whether the preset sleep condition is met by combining the wake-up threshold in the stored operating parameters, the valid time limit information, and the task planning;

[0093] S22: The auxiliary processing unit generates instructions containing data saving identifier, credibility, and time limit verification results, and task planning association description according to the determination result;

[0094] S23: The main processing unit receives the instruction, confirms the data validity and task planning matching according to the verification result, executes the instruction, and feeds back the result to the auxiliary processing unit.

[0095] In this embodiment, first, after the main processing unit completes a video collection upload, it transmits the running state data to the auxiliary processing unit: the task state marker is "completed", the environmental data contains the current temperature 25℃, the battery power 80%, the credibility identifier is set to 98% (indicating that the data is reliable), the time limit information indicates "valid within 30 minutes before the next task starts", and the task planning clearly indicates "to execute the next collection after 30 minutes". These data are transmitted through the SPI interface and form the basis for judgment with the running parameters stored in S1. Then, after the auxiliary processing unit receives the data, it first verifies the credibility identifier (98% ≥ preset credibility threshold 90%) to confirm that the data is credible; then it filters the valid time limit information (excludes expired historical data) and comprehensively judges in combination with the wake-up threshold (temperature 0-50℃, battery power ≥20%) stored in S1: the current temperature 25℃ is within the threshold range, the power 80% meets the requirement, and the remaining valid time length 30 minutes matches the next start interval of the task planning, and the judgment result is "satisfy the sleep condition". Subsequently, the auxiliary processing unit generates corresponding instructions: the data storage identifier clearly indicates that the "core layer (video index) must be stored to 0x0100 address, and the extended layer (environmental log) can be optionally stored to 0x0200 address"; the credibility and time limit verification results are marked as "data is credible and time limit is valid"; the task planning association indicates that the key node (next start time T+30 minutes) needs to be loaded preferentially. Finally, the main processing unit receives the instruction, checks that the verification result is consistent with the local task planning (confirms that it is correct to start after 30 minutes), executes the data storage operation, feeds back "storage is completed" to the auxiliary processing unit, prepares for cutting off the main power supply and entering sleep in S3, and ensures smooth connection of the entire process. Through multi-dimensional data verification and structured instruction generation, it is ensured that the sleep judgment is accurate and traceable, the task planning association guarantees the connection of the subsequent process, and the system collaboration efficiency and decision reliability are improved.

[0096] In an embodiment, the auxiliary processing unit verifies the credibility identifier in the running state data, and comprehensively judges whether the preset sleep condition is met in combination with the wake-up threshold, the valid time limit information, and the task planning stored in the running parameters, and the step comprises:

[0097] S210: Verify the credibility identifier in the running state data, filter the credible data, extract the key parameters and time stamp, check the data integrity and supplement the missing items;

[0098] S211: Compare the key parameters with the wake-up threshold in the stored running parameters, mark the abnormal items, exclude the expired abnormal items in combination with the valid time limit information, and calculate the remaining valid time length of the data;

[0099] S212: compare the remaining effective duration with the shortest hibernation duration, and refer to the next start time and resource requirement of the task planning;

[0100] S213: determine whether the abnormal item influence and resource support hibernation;

[0101] S214: determine whether the hibernation condition is met based on the above results, and generate a result containing the determination basis.

[0102] In this embodiment, first, the auxiliary processing unit receives the running state data transmitted by the main processing unit, and then verifies the credibility mark (98%), confirms that it is higher than the preset 90% credibility threshold, and filters as credible data; Then extract the key parameters (current temperature 25℃, battery power 80%) and time stamp (10:00:00), check and find that the "environmental humidity" item is missing, automatically supplement the historical average value 60% (based on similar scene data), and ensure the data integrity. Next, the auxiliary processing unit calls the wake-up threshold (temperature 0-50℃, battery power ≥20%) stored in S1, compares the key parameters with it: the temperature 25℃ is in the normal range, and the power 80% meets the requirement, and there is no abnormal item; Combined with the time limit information ("valid for 30 minutes before the next task"), the remaining effective duration of the data is calculated as 30 minutes (current 10:00 to next task 10:30), and all expired abnormal items are excluded (no in this case). After that, compare the remaining effective duration 30 minutes with the preset shortest hibernation duration 25 minutes, confirm that it meets the basic hibernation duration requirement; At the same time, refer to the next start time (10:30) and resource requirement (video acquisition power ≥15%) of the task planning, judge that the current power 80% can support hibernation and next task execution. Finally, the comprehensive judgment is: no abnormal item influence, sufficient resources, remaining effective duration covers hibernation period, meets hibernation condition, generates determination basis result containing "temperature normal, power sufficient, time limit matching, resource support", provides detailed logical support for subsequent instruction generation, and ensures that the hibernation decision of S2 completely adapts to the actual running state of the device. Through data completion, multi-dimensional comparison and resource evaluation, the fine hibernation condition is realized, avoiding misjudgment due to data loss or single factor, ensuring that the hibernation decision is scientific and reliable, and improving the accuracy of low-power control of the system.

[0103] In one embodiment, the auxiliary processing unit generates instructions containing data saving marks, credibility and time limit verification results, and task planning association instructions according to the determination results, including:

[0104] S220: generating data storage marks containing storage path, data type and storage priority according to the determination result, distinguishing between mandatory and optional data, wherein the mandatory data is the data necessary for the system to restore to the state before hibernation after hibernation, and the optional data is the auxiliary analysis or non-essential supplementary data, including environmental data details and non-core task logs;

[0105] S221: generating verification summary, status code and report fragment from the credibility and timeliness verification result;

[0106] S222: extracting key nodes in combination with task planning association description, generating association index and time node;

[0107] S223: integrating data storage marks, verification results and key node descriptions, adding integration verification information to form instructions.

[0108] In this embodiment, first, the auxiliary processing unit generates data storage marks according to the determination result of "S2 meets the hibernation condition": the storage path is 0x0100-0x0110 (core layer) and 0x0200-0x0205 (extension layer) of EEPROM, the data type is distinguished as "video index (binary)" and "environmental log (text)", the storage priority is marked as "video index P0 (mandatory)" and "environmental log P1 (optional)", and the key data is ensured to be stored by the main processing unit first. Then, for the credibility and timeliness verification result, the verification summary is "data credibility 98%, timeliness valid (remaining 30 minutes)", the status code is set to "0x00 (verification passed)", and the report fragment is excerpted as "temperature 25℃ within threshold range, power 80% meet demand", which is convenient for the main processing unit to quickly understand the verification conclusion without parsing the original data, wherein the mandatory data is the data necessary for the system to restore to the state before hibernation after hibernation, and the optional data is the auxiliary analysis or non-essential supplementary data, including environmental data details and non-core task logs.

[0109] Subsequently, in combination with the task planning association description, the key nodes are extracted: next start time (10:30), main task (video acquisition), subtask (4G upload), resource allocation (power consumption ≥ 15%); a hierarchical association index (main task ID: 001 binds subtask ID: 001-1) is generated, a state association index (conversion logic of "dormancy → wake up → running") is generated, and each node time node (such as the subtask needs to be completed before 10:32) is marked, to ensure that the main processing unit clearly connects the task after dormancy. Finally, the above information is integrated: according to the order of "data saving identifier → verification result → key node", the integrated verification information (CRC16 value 0x5678) is added to form a complete instruction. After receiving, the main processing unit confirms that the instruction is not tampered with by the verification value, stores data according to the identifier, refers to the node planning task, provides clear operation guidance for the S3 sleep execution, and ensures seamless connection with the subsequent process. Through the design of structured instructions, the data storage priority and task node are clear, the verification result is intuitive and easy to understand, the integrated verification ensures the integrity of the instruction, and the system collaboration efficiency and sleep execution accuracy are improved.

[0110] In one embodiment, the step of extracting key nodes, generating association indexes, and marking time nodes in combination with the task planning association description includes:

[0111] S2220: Analyze the task planning association description, and identify the task start, end, main task, subtask, resource allocation, and state conversion node as key nodes;

[0112] S2221: Assign a unique identifier to the key nodes, generate a hierarchical association index to mark the subordinate relationship, and generate a state association index to record the conversion logic;

[0113] S2222: Extract the execution time, planned execution point, and state conversion expected time of each node as time nodes and mark them;

[0114] S2223: Organize the key nodes, hierarchical association indexes, state association indexes, and time nodes into an association table according to the corresponding relationship;

[0115] S2224: Arrange the association table according to the subordinate relationship and state conversion logic.

[0116] In this embodiment, first, the auxiliary processing unit parses the task planning association description ("start video collection main task at 10:30, trigger 4G upload subtask after completion, required power ≥ 15%, state from hibernation → wake up → running"), identifies the key nodes: task start node (10:30 video collection begins), task end node (10:35 video collection is completed), main task node (video collection), subtask node (4G upload), resource allocation node (power ≥ 15%), and state transition node (hibernation → wake up, wake up → running). Next, unique identifiers are assigned to each key node: start node ID001, end node ID002, main task ID003, subtask ID004, resource node ID005, state transition node ID006 (hibernation → wake up), and ID007 (wake up → running). A hierarchical association index is generated: the "ID003 → ID004" label marks the subordinate relationship between the main task and the subtask, and the "ID005 → ID003" label marks the support relationship of resource allocation to the main task. A state association index is generated: "ID006 → ID007" records the logical sequence of state transitions, ensuring that the task execution order is clear.

[0117] Subsequently, the time nodes of each node are extracted: the start node execution time is 10:30, and the planned execution point is 10:30 (no deviation); the end node execution time is 10:35, and the planned execution point is 10:35; the state transition node ID006 expected time is 10:29 (wake up begins), and the ID007 expected time is 10:30 (enter running), and the corresponding annotations are made in the association table. Finally, the association table is sorted according to the "subordinate relationship (main → sub)" and the "state transition logic (hibernation → wake up → running)": first list the resource node ID005, then the main task ID003 and its subtask ID004, and finally the state transition nodes ID006, ID007 and the start / end nodes ID001, ID002, forming an ordered association table of "resource → main task → subtask → state transition → time node", which provides precise node connection basis for the main processing unit to resume the task after S3 hibernation and S4 wake up. Through the detailed analysis and association index construction of the key nodes, the task subordinate relationship and state transition logic are clear, the time node annotation ensures the timing accuracy, and clear guidance is provided for subsequent task recovery, improving the reliability of system process connection.

[0118] In one embodiment, the main processing unit, in response to the hibernation instruction, hierarchically writes the key data before hibernation to the non-volatile storage unit and feeds back the data identifier and the log, which are verified by the auxiliary processing unit. After verification, the auxiliary processing unit controls the power supply circuit of the main processing unit to be cut off, and the auxiliary processing unit enters a low-power state after binding the relevant data. The steps include:

[0119] S30: The main processing unit writes the key data before hibernation into the non-volatile storage unit in layers, generates a data identifier for each layer and a storage log, and feeds back the data identifier, log digest, and layered storage result to the auxiliary processing unit;

[0120] S31: The auxiliary processing unit checks the data identifier matching and the log digest integrity, and confirms the completion of the storage of each layer of data. After the completion of the storage of each layer of data, the auxiliary processing unit cuts off the power supply loop of the main processing unit;

[0121] S32: The auxiliary processing unit binds the data identifier, log digest, and main processing unit state parameter according to the data layer correspondence relationship.

[0122] S33: The auxiliary processing unit associates the data layer information with the binding record, and enters the low-power state carrying the associated record.

[0123] In this embodiment, in the industrial field 4G monitoring device, the sub-step fully presents the operation flow of hibernation execution, and is deeply connected with the instruction generated in S2. First, after the main processing unit receives the hibernation instruction sent by the auxiliary processing unit in S2, according to the requirement of "core layer data must be stored, and extension layer data can be stored", the key data is written into EEPROM in layers: the core layer data is video index (records the collection time 10:00, the storage address 0x0100-0x0110), which is written in binary format; the extension layer data is environment log (temperature 25℃, power 80%), which is written in text format in the address segment 0x0200-0x0205. After the writing is completed, the main processing unit generates a SHA-256 data identifier (0x1A2B3C4D) for the core layer, a data identifier (0x5E6F7G8H) for the extension layer, and a storage log (records the writing time 10:02, the address range, the data size), and extracts a log digest ("core layer 16 bytes, extension layer 8 bytes, writing no exception"), and feeds back these information to the auxiliary processing unit through SPI.

[0124] Then, after receiving the feedback, the auxiliary processing unit first checks the matching of the data identifier: the preset core layer data identifier generation rule in S1 is called, the SHA-256 value of the video index is recalculated, and it is confirmed that it is consistent with 0x1A2B3C4D fed back by the main processing unit; then the log summary integrity is verified, the address range and data size in the summary are checked to see whether they match the actual storage situation, and after confirming that there is no loss or error, a "power cut permission" signal is generated to trigger the MOS tube to cut off the main power supply circuit of the main processing unit, at this time the main processing unit is completely powered off, the power consumption is reduced to 0 muA, and the standby power consumption of the core module in the traditional sleep mode is avoided. Subsequently, the auxiliary processing unit binds the core layer data identifier (0x1A2B3C4D), the core layer log summary, and the state parameters before the main processing unit sleeps ("task completed, power 80%, sleep trigger source is timing") according to the data layering correspondence, generates binding group 1; binds the extension layer data identifier (0x5E6F7G8H), the extension layer log summary, and the corresponding state parameters, generates binding group 2, and adds the layering labels of "core layer" and "extension layer" to the two binding groups respectively.

[0125] Finally, the auxiliary processing unit associates the data layering information (address range, data type) with the two binding groups, generates an association record (containing "binding group 1 corresponds to 0x0100-0x0110 core data" and "binding group 2 corresponds to 0x0200-0x0205 extension data"), adds a CRC check code (0x9I0J1K2L) to the association record, carries the association record into the low-power standby mode (power consumption is only 50 muA), and provides complete data association basis for the main processing unit to quickly locate and read the data before sleep and recover the state in the subsequent S4, ensuring the process closed loop. Through layered storage and data identifier verification, it is ensured that the key data is not lost or tampered with; the power supply of the main unit is completely cut off to maximize the reduction of power consumption, and the binding record and association information lay the foundation for wake-up recovery, taking into account the low-power effect and system reliability.

[0126] In one embodiment, the auxiliary processing unit checks the data identifier matching and the log summary integrity, and after confirming that the data storage of each layer is completed, the step of cutting off the power supply circuit of the main processing unit by the auxiliary processing unit includes:

[0127] S310: Receive the data identifier, the log summary, and the layered storage result, distinguish the core layer and the extension layer, sort by level and assign verification serial numbers;

[0128] S311: Verify the core layer data identifier matching and the log summary format compliance in order of serial number, then verify the extension layer, and record the verification result;

[0129] S312: After confirming that all levels pass the verification and the summary is complete, a power cut permission signal is generated;

[0130] S313: Cut off the power supply loop of the main processing unit according to the permission signal.

[0131] In this embodiment, first, the auxiliary processing unit receives the information fed back by the main processing unit: the core layer data identifier (SHA-256 value 0x1A2B3C4D), the extended layer data identifier (0x5E6F7G8H), the core layer log digest (“10:02 write 0x0100-0x0110, 16 bytes, no exception”), the extended layer log digest (“10:02 write 0x0200-0x0205, 8 bytes, no exception”), and the hierarchical storage result (core layer complete, extended layer complete). According to the “core layer priority” rule preset in S1, the auxiliary processing unit distinguishes the core layer and the extended layer, sorts them in the order of “core layer→extended layer”, allocates the verification serial number “001” to the core layer and “002” to the extended layer, and ensures that the key data is verified first. Then, the auxiliary processing unit starts verification according to the serial number: first, verify the core layer with serial number “001” — call the core layer data identifier generation algorithm (SHA-256) stored in S1, re-read the video index data in the EEPROM 0x0100-0x0110 address segment, calculate the SHA-256 value, and find that it completely matches 0x1A2B3C4D fed back by the main processing unit, confirming that the data identifier matches; at the same time, check the core layer log digest format, confirm that it conforms to the preset format “write time+address range+data size+state” (such as “10:02 write 0x0100-0x0110, 16 bytes, no exception” completely meets the format requirements), and determine that the log digest format is compliant. After completing the core layer verification, verify the extended layer with serial number “002”: using the same logic, re-calculate the data identifier of the environmental log in the 0x0200-0x0205 address segment, which is consistent with 0x5E6F7G8H, and the log digest format also meets the requirements, record both verification results as “pass” and store them to the temporary buffer area.

[0132] Subsequently, the auxiliary processing unit aggregates the verification results: core layer data identification matches, log compliance, extended layer data identification matches, log compliance, all levels pass verification, and there is no data missing, address error, etc. in the log summary, confirming the completeness of the summary. Based on this, the auxiliary processing unit generates a power cut permission signal, which contains the "verification passed" status code (0x00), verification time (10:03) and target power supply loop identification (main processing unit main power supply loop), ensuring the uniqueness of the signal. Finally, the auxiliary processing unit sends a trigger instruction to the MOS tube controller according to the permission signal, and the MOS tube turns on the cut-off loop after receiving the instruction, completely disconnecting the main power supply of the main processing unit (from the previous 1.8V power supply to 0V), at which time the main processing unit stops all unnecessary operations, only retaining the data of the non-volatile storage unit, preparing for data reading and state recovery when S4 wakes up later, while maximizing the power consumption during hibernation. Through hierarchical sorting and priority verification, the core data storage is guaranteed to be reliable; full-level verification and format checking avoid risks caused by data or log abnormalities, and the permission signal mechanism ensures that the power cut operation is rigorous, improving the safety and stability of system hibernation execution.

[0133] In an embodiment, the steps of receiving data identification, log summary and hierarchical storage results, distinguishing core layer and extended layer, sorting by level and assigning verification sequence number include:

[0134] S3101: Receive data identification, log summary and hierarchical storage results, identify each layer type identification and non-replaceable identification, extract the core layer containing non-replaceable identification and corresponding type, and the rest as extended layer;

[0135] S3102: Extract the storage time stamp and capacity information of the core layer and the extended layer, and sort them by time stamp and capacity from large to small;

[0136] S3103: Assign each layer after sorting a verification sequence number containing a time stamp segment and a capacity proportion, associate each layer type identification, sorting result and verification sequence number to form a hierarchical verification sequence.

[0137] In this embodiment, first, the auxiliary processing unit receives the information fed back by the main processing unit: core layer data identifier (SHA-256 value 0x1A2B3C4D), extended layer data identifier (0x5E6F7G8H), core layer log summary (containing type identifier "VIDEO_INDEX", non-substitutable identifier "TRUE"), extended layer log summary (containing type identifier "ENV_LOG", non-substitutable identifier "FALSE"), and hierarchical storage result (core layer and extended layer both complete storage). The auxiliary processing unit extracts the "VIDEO_INDEX" type data containing "non-substitutable identifier = TRUE" as the core layer (corresponding to the video index, which is the key basis for resuming the task after S4 wakes up) and the "ENV_LOG" type data with "non-substitutable identifier = FALSE" as the extended layer (corresponding to the environment log, only for auxiliary reference) by analyzing the identifiers in the log summary, and clearly distinguishes the core degree difference between the two layers. Then, the auxiliary processing unit extracts the storage timestamp and capacity information from the log summary: the core layer storage timestamp is 10:02:05, and the data capacity is 16 bytes; the extended layer storage timestamp is 10:02:08, and the data capacity is 8 bytes. According to the rule of "timestamp first + capacity from large to small", the core layer timestamp is earlier (10:02:05 is earlier than 10:02:08), and the capacity is larger (16 bytes is larger than 8 bytes), so the sorting result is "core layer → extended layer", which ensures that the key data with earlier storage time and larger size is processed first.

[0138] Finally, the auxiliary processing unit assigns verification serial numbers to the sorted two-layer data: the timestamp segment of the core layer is extracted as "100205", the capacity proportion (the total capacity of the two layers is 24 bytes, and the core layer proportion is 66.7%) is marked as "66.7%", and the verification serial number "100205-66.7%-001" (the last "001" is the sorting serial number) is generated; the extended layer timestamp segment is "100208", the capacity proportion is 33.3%, and the verification serial number "100208-33.3%-002" is generated. Then, the type identifier (core layer "VIDEO_INDEX", extended layer "ENV_LOG"), the sorting result (core layer first, extended layer second), and the verification serial number are associated to form a hierarchical verification sequence ("serial number 100205-66.7%-001: type VIDEO_INDEX, sorting 1; serial number 100208-33.3%-002: type ENV_LOG, sorting 2"), which lays a clear execution framework for subsequent verification of the core layer in order of serial number, ensuring seamless connection with the subsequent verification steps of S3. By accurately distinguishing the core and extended layers through identification, prioritizing the processing of key data according to timestamp and capacity, and using verification serial numbers with features for traceability, the subsequent verification provides an ordered basis, improving the accuracy and efficiency of data processing.

[0139] In one embodiment, the auxiliary processing unit binds the data identification, the log summary, and the main processing unit state parameter record in a data layer correspondence relationship, including:

[0140] S320: Identifying the data layer identification, distinguishing the core layer and the extension layer, extracting the data identification feature information and the log summary key information of each layer;

[0141] S321: Generating a merged check value for the core layer and the extension layer respectively, binding the feature information, the key information, the check value, and the corresponding layer main processing unit state parameter, and adding a layer type label;

[0142] S322: Adding a layer identification as an association key to each binding group to establish an association record of the type label and the binding group;

[0143] S323: Storing the association record and all binding groups in order of the layer identification.

[0144] In this embodiment, first, the auxiliary processing unit identifies the data layer identification - the layer storage result and the log summary fed back from the main processing unit in S3, extracts the core layer identification "VIDEO_INDEX-CORE" and the extension layer identification "ENV_LOG-EXT", and clearly distinguishes the two layers of data; then extracts the key information of each layer: the core layer data identification feature information is the first 8 bits "1A2B3C4D" of the SHA-256 value 0x1A2B3C4D, and the log summary key information is "10:02 Write 0x0100-0x0110, 16 bytes"; the extension layer data identification feature information is the first 8 bits "5E6F7G8H" of 0x5E6F7G8H, and the log summary key information is "10:02 Write 0x0200-0x0205, 8 bytes", ensuring that the extracted information can uniquely correspond to each layer of data. Subsequently, the auxiliary processing unit generates a merged check value for the two layers of data respectively: for the core layer feature information, key information, and main processing unit state parameter ("task completed, power 80%, hibernate trigger source timing"), the merged check value 0x98765432 is calculated using the CRC32 algorithm; for the feature information, the key information, and the corresponding state parameter ("environmental temperature 25°C, humidity 60%") of the extension layer, the merged check value 0x12345678 is calculated. The core layer feature information, key information, and merged check value are bound with the state parameter, and a layer type label "CORE-LAYER" is added; similarly, the extension layer binding information is labeled "EXT-LAYER", forming two independent binding groups, ensuring that each layer of data corresponds to the state parameter one by one.

[0145] After that, the auxiliary processing unit adds a hierarchical identifier as an association key for each binding group: the association key of the core layer binding group is "VIDEO_INDEX-CORE-0x0100" (containing the hierarchical identifier and the starting address), and the association key of the extension layer binding group is "ENV_LOG-EXT-0x0200"; by comparing the hierarchical identifier in the key with the hierarchical type label, the association record of the "CORE-LAYER" label and the core layer binding group and the "EXT-LAYER" label and the extension layer binding group is established, and the correspondence between the key and the binding group is clearly marked in the record to avoid confusion. Finally, the auxiliary processing unit stores the association record and the binding group in order of the hierarchical identifier (core layer first): first, the core layer association record and the binding group are stored in the 0x0300-0x0320 address segment of the EEPROM, and then the extension layer related information is stored in the 0x0321-0x0340 address segment, and a storage timestamp 10:04 is attached when storing to ensure that when the S4 is woken up later, the corresponding binding group can be quickly located and read according to the hierarchical identifier, providing complete association basis for data recovery, realizing process closed loop. Through hierarchical identifier and key information extraction, the data association is accurate; the combination of the check value and the association key ensures the data integrity and traceability, and the ordered storage speeds up the subsequent wake-up recovery, improving the cooperative reliability of system sleep and wake-up.

[0146] In one embodiment, the step of adding a hierarchical identifier as an association key for each binding group and establishing an association record of the type label and the binding group comprises:

[0147] S3220: A unique hierarchical identifier containing hierarchical information and a check bit is generated for each binding group as an association key embedded in the binding group;

[0148] S3221: Identify the attribute features of the type label and classify, and extract features from the classified label;

[0149] S3222: By comparing the association key features and the label features, the type label is associated with the corresponding binding group, and an association record is generated;

[0150] S3223: Add a key check code to the association record, and store the correspondence table of the label features and the association record.

[0151] In this embodiment, first, the auxiliary processing unit generates a unique hierarchical identifier for the core layer binding group: the fusion level information "CORE-001" ("CORE" represents the core layer, and "001" is the serial number) and the CRC8 check bit "0xAB" are combined as "CORE-001-AB" as the association key, and the core layer binding group (including video index data identifier, state parameters, etc.) is embedded; generate "EXT-002-CD" (level information "EXT-002" + check bit "0xCD") as the association key for the extension layer binding group, and embed the extension layer binding group to ensure that the key is unique and verifiable. Next, identify the attribute characteristics of the type label: the attribute characteristics of the core layer label "CORE-LAYER" are "non-replaceable, mission critical", and the attribute characteristics of the extension layer label "EXT-LAYER" are "replaceable, auxiliary reference". After classification according to "criticality", the core layer label feature code "KEY-CORE" and the extension layer label feature code "KEY-EXT" are extracted, and the features are simplified for comparison.

[0152] Subsequently, the association key features are compared with the label features: the feature field "CORE" of the core layer key "CORE-001-AB" matches the label feature code "KEY-CORE", and the feature field "EXT" of the extension layer key "EXT-002-CD" matches "KEY-EXT". The type label is successfully associated with the corresponding binding group, and an association record is generated ("CORE-LAYER → CORE-001-AB binding group; EXT-LAYER → EXT-002-CD binding group"). Finally, a key check code is added to the association record (a CRC16 value "0xEF12" is calculated for the record content), and a corresponding table of label features and association records ("KEY-CORE → Association Record 1; KEY-EXT → Association Record 2") is stored in the 0x0341 address segment of the EEPROM, ensuring that the binding group can be quickly located by the label and the key integrity can be verified when the S4 is woken up in the future. The unique key and feature comparison ensure accurate association of the label and the binding group, the check code ensures record integrity, the corresponding table improves data positioning efficiency when waking up, and the security and traceability of system data association are enhanced.

[0153] In one embodiment, the attribute characteristics of the type label are identified and classified, and the step of feature extraction for the classified labels includes:

[0154] S32210: Scan the type label, extract the basic attributes, dynamic and static attributes, and association attribute characteristics, determine the association object and the association strength;

[0155] S32211: Classify according to the association strength, combine the dynamic and static attributes, and form an association-dynamic and static two-dimensional classification matrix;

[0156] S32212: Extract common features, differentiated features and correlation feature coefficients for the same type of labels in the classification matrix.

[0157] S32213: Associate the hierarchical classification results with the feature extraction data, and bind the two-dimensional classification and feature coefficients.

[0158] In this embodiment, first, the auxiliary processing unit scans the type labels "CORE-LAYER" and "EXT-LAYER": extracts the basic attribute ("core layer data identifier"), static attribute ("non-replaceable"), dynamic attribute ("updated with task period"), and correlation attribute ("bound to video index data") of "CORE-LAYER", determines the correlation object as the video index storage module, and the correlation strength is marked as "high (90%)"; extracts the basic attribute ("extension layer data identifier"), static attribute ("replaceable"), dynamic attribute ("updated on demand"), and correlation attribute ("bound to environment log data") of "EXT-LAYER", and the correlation object is the environment sensor module, and the correlation strength is marked as "medium (60%)". Then, according to the correlation strength classification (high / medium) and the dynamic / static attribute division, a two-dimensional classification matrix is formed: "CORE-LAYER" falls into the "high correlation strength + dynamic attribute" quadrant, and "EXT-LAYER" falls into the "medium correlation strength + static attribute" quadrant, which clearly distinguishes the core degree and update characteristics of the labels through the matrix.

[0159] Subsequently, the features of the same type of labels (two independent labels in this case) in the matrix are extracted: the common feature of "CORE-LAYER" is "strongly related to task recovery", the differentiated feature is "need to verify integrity in real time", and the correlation feature coefficient (binding degree with video index) is set to 0.9; the common feature of "EXT-LAYER" is "auxiliary environment analysis", the differentiated feature is "lack of impact on main process", and the correlation feature coefficient (binding degree with environment log) is set to 0.6. Finally, the hierarchical classification results ("CORE-LAYER: high correlation-dynamic" "EXT-LAYER: medium correlation-static") are associated with the feature extraction data, the two-dimensional classification and feature coefficients are bound, and a feature comparison table is generated and stored in EEPROM, which provides a basis for subsequent S4 to quickly identify label attributes and prioritize high correlation data, ensuring consistency with the overall process loop. Through multi-dimensional attribute extraction and matrix classification, the importance and characteristics of the labels are accurately distinguished; the correlation strength is quantified, which provides a basis for data processing priority and improves the refinement level of system data management.

[0160] In one embodiment, the auxiliary processing unit associates data layer information with binding records, and the step of carrying the association records into a low-power state includes:

[0161] S330: Extract the level identifier, timestamp and priority in the data layer information, and extract the association key, time marker and priority label in the binding record;

[0162] S331: Generate an association index according to the matching of the level identifier and the association key, generate a time sequence association table according to the association of the timestamp and the time marker, and generate an association queue according to the correspondence of the priority and the priority label;

[0163] S332: Add integrity check identifier and time validity identifier to the association index, the time sequence association table and the association queue;

[0164] S333: Carry the association index, the time sequence association table and the association queue with the integrity check identifier and the time validity identifier into the low-power state.

[0165] In this embodiment, first, the auxiliary processing unit extracts the data layer information: the core layer level identifier "VIDEO_INDEX-CORE", the timestamp 10:02:05, and the priority "P0 (highest)"; the extension layer level identifier "ENV_LOG-EXT", the timestamp 10:02:08, and the priority "P1 (second highest)". At the same time, the binding record information is extracted: the core layer association key "CORE-001-AB", the time marker 10:04:00, and the priority label "CRITICAL"; the extension layer association key "EXT-002-CD", the time marker 10:04:05, and the priority label "SUPPLEMENTARY", to ensure that the extracted information covers the three key dimensions of level, time and priority. Then, the association key is matched according to the level identifier: "VIDEO_INDEX-CORE" is matched with "CORE-001-AB" to generate the association index "INDEX-CORE: VIDEO_INDEX-CORE→CORE-001-AB"; "ENV_LOG-EXT" is matched with "EXT-002-CD" to generate "INDEX-EXT: ENV_LOG-EXT→EXT-002-CD". The time marker is associated according to the timestamp: the core layer 10:02:05 (storage time) is associated with 10:04:00 (binding time), and the extension layer 10:02:08 is associated with 10:04:05 to generate the time sequence association table "[CORE: 10:02:05→10:04:00; EXT: 10:02:08→10:04:05]". The priority label is generated according to the correspondence of the priority: "P0" corresponds to "CRITICAL", and "P1" corresponds to "SUPPLEMENTARY", to generate the association queue "[CRITICAL (P0)→SUPPLEMENTARY (P1)]", to clearly indicate the data processing order.

[0166] Subsequently, the integrity check identifier (CRC32 value "0x11223344" is calculated for the content) and the time validity identifier ("VALID_UNTIL: 10:30:00", which matches the next wake-up time) are added to the association index, the time sequence association table, and the association queue to ensure that the association information is not tampered with during hibernation and is valid within the time limit. Finally, the auxiliary processing unit packages the association index, the time sequence association table, and the association queue with the check identifier and the validity identifier into the internal RAM low-power area, cuts off the power supply of unnecessary modules, enters the low-power state with a power consumption of 50 μA, and waits for the wake-up event trigger in S4 to ensure that the association information can be quickly called to restore the system when it wakes up. Through multi-dimensional association integration, a complete data association system is constructed, the check and time limit identifier ensures information reliability, the low-power state carries information to speed up the recovery of wake-up, and both low power consumption and data traceability are considered.

[0167] In an embodiment, after the auxiliary processing unit monitors the wake-up event and verifies it, the auxiliary processing unit controls the main processing unit power supply loop to be turned on, transmits the wake-up information and the data storage location to the main processing unit, and the main processing unit reads the key data before hibernation to restore the state before hibernation. The steps include:

[0168] S40: The auxiliary processing unit monitors the wake-up event related signals and data, and starts the wake-up process when the trigger condition is met, records the trigger time and extracts the trigger feature;

[0169] S41: The auxiliary processing unit compares the trigger time with the wake-up permission period and matches the trigger feature with the wake-up feature library, and turns on the main processing unit power supply loop after double verification;

[0170] S42: The auxiliary processing unit transmits the trigger time, the feature matching confidence value, and the storage path of the key data before hibernation to the main processing unit;

[0171] S43: The main processing unit reads the corresponding hierarchical data and state parameters according to the storage path, and restores to the state before hibernation according to the parameters.

[0172] In this embodiment, first, the auxiliary processing unit continuously monitors the wake-up event initialized in S1 in a low-power state: when the 30-minute timer triggers (reaches the 10:30 wake-up time planned in S2), the auxiliary processing unit immediately records the trigger time as 10:30:00, extracts the trigger feature as "timer wake-up (period 30 minutes)", and checks whether there are other wake-up events (such as temperature sensor triggering) at the same time, confirming that the source of this wake-up is the timer. Next, the auxiliary processing unit performs double verification: compares the trigger time 10:30:00 with the "wake-up permission period" (00:00-24:00) stored in S1, confirming that it is within the permitted range; calls the pre-set "wake-up feature library" in S1, compares the "timer wake-up (period 30 minutes)" feature with the "timer trigger template" in the library, and the matching degree reaches 100%, passing the double verification. Subsequently, the auxiliary processing unit controls the MOS tube to turn on the main processing unit power supply circuit, and the main processing unit recovers power supply (voltage rises to 1.8V) from the power-off state, and starts to initialize.

[0173] Subsequently, the auxiliary processing unit transmits the wake-up information to the main processing unit: the trigger time 10:30:00, the feature matching confidence value "100% match timer template", and the key data storage path recorded in S3-the core layer data (video index) is located at EEPROM 0x0100-0x0110, and the extended layer data (environment log) is located at 0x0200-0x0205, with data identification (core layer 0x1A2B3C4D, extended layer 0x5E6F7G8H) attached for verification. Finally, after receiving the information, the main processing unit reads the layered data according to the storage path: first reads the core layer video index, and then reads the extended layer environment log through data identification verification to confirm that it has not been tampered with; at the same time, the state parameters bound to the record in S3 ("task completed, power 80%") are called to synchronize the current state to the task completion node before hibernation, load the next task planning (start video collection at 10:30), complete state recovery, and prepare for the execution of new tasks, forming a complete closed loop with the task planning in S2 and the hibernation storage in S3. Double verification avoids false wake-up to ensure system stability, accurate transmission of storage path and data identification ensures recovery efficiency, and seamless state connection ensures task continuity, improving system wake-up reliability and response speed.

[0174] In one embodiment, the step of the auxiliary processing unit comparing the trigger time with the wake-up permission period, matching the trigger feature with the wake-up feature library, and passing the double verification includes:

[0175] S410: The auxiliary processing unit extracts the trigger time and the trigger feature, divides the wake-up permission period and the feature library by level, adds a timestamp to the trigger time, and generates a feature code for the trigger feature;

[0176] S411: The auxiliary processing unit compares the timestamp with the corresponding level permission period, compares the feature code with the corresponding level feature library, and generates a feature matching confidence value of both;

[0177] S412: The auxiliary processing unit associates the period level and the feature level, verifies the consistency and whether the feature matching confidence value meets the standard;

[0178] S413: When the auxiliary processing unit level is consistent and the feature matching confidence value meets the standard, a power-on signal is generated to turn on the power supply loop of the main processing unit.

[0179] In this embodiment, first, the auxiliary processing unit extracts the trigger time 10:30:00 and the trigger feature "timed wake-up (period 30 minutes)", divides the wake-up permission period level (core layer: 08:00-20:00, extended layer: 20:00-08:00) and the feature library level (core library: timed trigger template, extended library: temperature abnormal template) according to S1 preset rules. Add a millisecond timestamp "10:30:00.123" to the trigger time, generate a feature code "TS-30MIN-789" for the trigger feature through a hash algorithm, and ensure the unique identification of the time and the feature. Then, compare the timestamp with the corresponding level permission period: 10:30:00.123 falls within the core layer permission period 08:00-20:00, and generate a period feature matching confidence value of 95% (because it completely matches the core period); compare the feature code "TS-30MIN-789" with "timed trigger template" in the core layer feature library, and the feature field completely matches, generating a feature matching confidence value of 100%. Then, associate the period level and the feature level: confirm that the timestamp belongs to the core layer and the core layer matched by the feature code is consistent, without level crossing conflict; at the same time, verify that both feature matching confidence values (95%≥80% threshold, 100%≥80% threshold) meet the standard, meeting the double verification condition. Finally, the auxiliary processing unit generates a power-on signal containing the level identifier "CORE" and the check code "0xAB12", and sends it to the MOS tube controller. After the controller verifies that the signal is valid, the main processing unit power supply loop is turned on, and the main processing unit starts to power on, laying a foundation for subsequent data reading and state recovery, and forming a closed loop with the S4 overall wake-up process. The level division and the double feature matching confidence value verification improve the wake-up accuracy, the level consistency check avoids logical conflict, the power-on signal verification ensures operation safety, effectively prevents false wake-up and accelerates the wake-up response.

[0180] In one embodiment, the steps of extracting the trigger time and the trigger feature, dividing the wake-up permission period and the feature library by level, adding a timestamp to the trigger time, and generating a feature code for the trigger feature include:

[0181] S4101: Identify the trigger source type and the occurrence order, extract the corresponding trigger time and feature details, and mark the trigger priority in combination with the environmental parameters;

[0182] S4102: Divide the wake-up permission period into core layer and extended layer according to importance level, and divide the feature library according to type matching degree level, and synchronize the order and priority;

[0183] S4103: Add environmental timestamp containing layer identification, type and order to trigger time, and generate matching degree feature code for trigger feature;

[0184] S4104: Bind the time stamp level and the feature code level, associate the environmental parameters and the priority, and form the hierarchical matching set.

[0185] In this embodiment, this sub-step is a key information preprocessing link before wake-up verification, providing structured data support for subsequent double verification, closely linked with the wake-up event monitoring logic in the foregoing. First, the auxiliary processing unit identifies the trigger source type as "timer trigger" (from the 30-minute timer initialized in S1), the occurrence order is marked as "1" (the only trigger source this time), extracts the trigger time 10:30:00 and the feature details "period 30 minutes, no external interruption superposition", and marks the trigger priority as "high" (because it belongs to the wake-up in the preset task planning) in combination with the current environmental parameters (temperature 25℃, power 78%, both in the normal range). Then, the wake-up permission period is divided according to the importance level: the core layer period is "08:00-20:00" (the production peak period, which requires high-frequency wake-up monitoring), and the extended layer period is "20:00-08:00" (non-peak period, reduce the wake-up frequency); the feature library is divided according to the type matching degree level: the core library contains high-priority templates such as "timer trigger (30-minute period)" and "temperature threshold value (> 50℃)", and the extended library contains low-priority templates such as "humidity anomaly" and "weak signal", and the trigger order "1" and the priority "high" are synchronized and associated to ensure that the hierarchical division matches the trigger characteristics.

[0186] Subsequently, an environmental timestamp is added for the triggering time: the integration level identifier "CORE" (because 10:30 belongs to the core period), the triggering type "TIMER", the occurrence order "1", and the environmental parameters "T=25℃, P=78%", generate the timestamp "CORE-TIMER-1-10:30:00-T25-P78"; the triggering features "period 30 minutes, no external interruption" are matched with the "timing triggering template" in the core library (100%), and the feature code "MATCH-100%-TS30" is generated, which intuitively reflects the matching quality. Finally, the timestamp level "CORE" and the feature code level "CORE" (because the feature code corresponds to the core library template) are bound, the environmental parameters (normal) and the priority (high) are associated, and the level matching set "[Level: CORE; Timestamp: CORE-TIMER-1-10:30:00-T25-P78; Feature Code: MATCH-100%-TS30; Priority: High]" is formed, which provides a unified benchmark for subsequent level consistency verification and confidence calculation, and ensures seamless connection with the S4 wake-up verification process. Through level division and multi-dimensional information binding, the triggering information is structured and traceable; the environmental timestamp and the matching degree feature code improve the verification accuracy, the level matching set lays a foundation for subsequent verification, and the rigor of the wake-up logic is enhanced.

[0187] In one embodiment, the step of transmitting the triggering time, the feature matching confidence value, and the storage path of the key data before hibernation from the auxiliary processing unit to the main processing unit comprises:

[0188] S420: integrating the triggering time, the feature matching confidence value, and the storage path of the key data before hibernation, binding the triggering time and the storage path, adding a confidence label for the feature matching confidence value and associating the path index;

[0189] S421: extracting the triggering time slice, the feature slice corresponding to the feature matching confidence value, and splicing the storage path into composite information, adding a splicing identifier and a transmission verification code;

[0190] S422: generating a mapping set of binding relationships, associated indexes, and splicing identifiers;

[0191] S423: transmitting the composite information containing the transmission verification code and the mapping set to the main processing unit, so that the main processing unit verifies the verification code and analyzes the information according to the mapping set.

[0192] In this embodiment, first, the auxiliary processing unit integrates three types of core information: the trigger time "10:30:00.123" (including the timestamp of the double verification pass in S4), the feature match configuration confidence value "100% match between the timing trigger feature and the core library template", and the key data storage path before hibernation (core layer 0x0100-0x0110, extension layer 0x0200-0x0205, derived from the S3 hierarchical storage record). The trigger time is bound to the storage path as "10:30:00.123→[0x0100-0x0110, 0x0200-0x0205]", a confidence label "trusted (100%)" is added to the feature match configuration confidence value, and the path index "0x0100→CORE, 0x0200→EXT" is associated, ensuring clear correspondence between the information. Next, extract the trigger time slice "103000" (simplified time division seconds), the feature fragment "MATCH100" corresponding to the feature match configuration confidence value, which contains the confidence core value (such as 100%, 95%) and the match type identifier (such as timing trigger, threshold trigger), and the storage path "0x0100-0x0110; 0x0200-0x0205" is spliced into a composite information "103000-MATCH100-0x0100-0x0110; 0x0200-0x0205" in the order of "time slice + feature fragment + path", with a splicing identifier "SYNC-001" (indicating the first frame of synchronization information) and a transmission check code (calculating the CRC16 value "0x3456" for the composite information) added to prevent data tampering or loss during transmission.

[0193] Subsequently, a mapping set is generated: the binding relationship mapping "10:30:00.123 ↔ storage path", the index association mapping "trusted (100%) ↔ 0x0100→CORE", and the splicing identifier mapping "SYNC-001 ↔ composite information structure" are explicitly bound. These mapping relationships are arranged as a structured data table to facilitate fast parsing by the main processing unit. Finally, the auxiliary processing unit transmits the composite information containing the transmission check code and the mapping set to the main processing unit through the SPI interface. After receiving, the main processing unit first verifies that the check code "0x3456" is consistent with the local calculation result, and then parses the trigger time, feature match configuration confidence value, and storage path according to the mapping set, confirming that it completely matches the hierarchical storage location recorded in S3, preparing for subsequent data reading and state recovery, ensuring that the overall wake-up process of S4 and the S3 storage logic form a closed loop. Through information binding and composite splicing, transmission efficiency is improved, and check code and mapping set ensure information integrity and analyzability, precise association of wake-up events and storage data provides reliable basis for state recovery, and enhances the collaborative reliability of the main and auxiliary units.

[0194] In one embodiment, the step of splicing the feature segment corresponding to the feature matching confidence value, the time slice of the extraction trigger moment, and the storage path into a composite information, adding a splicing identifier and a transmission check code, comprises:

[0195] S4210: intercepting the key timestamp segment of the trigger moment, the feature hash segment of the feature matching confidence value, and the permission and version segment of the storage path;

[0196] S4211: in the order of timestamp-feature hash-permission-version-path, splicing to generate a composite information using a preset delimiter;

[0197] S4212: adding a splicing identifier containing the splicing order and the delimiter type to the composite information;

[0198] S4213: calculating the overall transmission check code based on the content of each segment, and adding the check code to the composite information.

[0199] In this embodiment, first, the auxiliary processing unit intercepts the key timestamp segment of the trigger moment "10:30:00.123" - "1030" (representing 10:30) with minute-level precision, taking into account timeliness and transmission efficiency; extracts the feature hash segment of the feature matching confidence value "100% matching degree of timing trigger feature and core library template", calculates the hash value of the matching result by SHA-1 algorithm, and intercepts the first 8 bits "F3A7B2C9" as the feature hash segment, simplifying the feature while retaining its uniqueness; extracts the permission and version segments from the storage path: the permission identifier of the core layer path 0x0100-0x0110 is "RW" (read-write), and the version number is "V1.0"; the permission identifier of the extension layer path 0x0200-0x0205 is "RO" (read-only), and the version number is "V1.0"; combined as "RW-V1.0; RO-V1.0", the data operation permission and version consistency are clear. Then, splice the information according to the preset rule: start with the timestamp segment "1030", sequentially splice the feature hash segment "F3A7B2C9", the permission and version segment "RW-V1.0; RO-V1.0", and the complete storage path "0x0100-0x0110; 0x0200-0x0205", separated by the delimiter "|", to generate the composite information "1030|F3A7B2C9|RW-V1.0; RO-V1.0|0x0100-0x0110; 0x0200-0x0205", ensuring that the information structure is clear and the boundaries of each segment are clear, facilitating the main processing unit to parse.

[0200] Subsequently, a splicing identifier is added to the composite information: the identifier contains the splicing sequence "TS→HASH→PERM→PATH" (timestamp→hash→permission→path) and the separator type "|", combined as "SYNC-ID-001|SEQ:TS→HASH→PERM→PATH|SEP:|", wherein "SYNC-ID-001" is a unique identifier, and the main processing unit can quickly confirm the information structure rule through the identifier to avoid parsing errors. Finally, the overall transmission check code is calculated based on the content of each segment: the CRC32 value is calculated for the timestamp segment, the feature hash segment, the permission version segment, and the storage path, and the four CRC values are subjected to secondary hash operation to obtain the final check code "0x7890ABCD", which is appended to the end of the composite information to form the complete transmission data "1030|F3A7B2C9|RW-V1.0;RO-V1.0|0x0100-0x0110;0x0200-0x0205|0x7890ABCD". After receiving, the main processing unit can verify whether the information is complete through the check code, forming a closed loop with the information transmission and subsequent data reading steps in S4, ensuring the accuracy of data recovery after wake-up. Through the segmentation processing of the transmission data, the structured splicing guarantees the orderliness of the information, and the splicing identifier and the check code improve the parsing efficiency and security, providing a reliable information basis for accurate data recovery after wake-up.

[0201] In one embodiment, the main processing unit extracts features based on historical data of multiple sleep-wake cycles, calculates an optimization coefficient to generate an update set, synchronizes the update set to the auxiliary processing unit, and the auxiliary processing unit updates the stored operating parameters and feeds back to the main processing unit. The steps include:

[0202] S50: distinguish core and extended historical data, based on historical data of multiple sleep-wake cycles, extract data fluctuation features and abnormal points, record time sequence;

[0203] S51: compare the core and extended data, fluctuation features and abnormal points with the corresponding reference respectively, calculate the combined optimization coefficient and abnormal correction coefficient, and associate the time stamp;

[0204] S52: add time effectiveness label and abnormal type identifier to the optimization coefficient to generate a hierarchical update set with time sequence marker;

[0205] S53: transmit the update set to the auxiliary processing unit, and the auxiliary processing unit updates the stored operating parameters according to the identifier classification, and feeds back the update result and abnormal processing situation to the main processing unit.

[0206] In this embodiment, first, the main processing unit distinguishes core historical data (such as 30 hibernation-wakeup cycle wakeup response time, core layer data verification pass rate) and extended historical data (environment log storage occupancy rate, extended layer data reading time consumption). Statistics show that: the wakeup response time fluctuates between 2.1-2.3 seconds (average 2.2 seconds), the core layer verification pass rate is stable at 100%, the environment log storage occupancy rate fluctuates with temperature rise (15% at 30°C, up to 25% at 45°C), and the extended layer reading time consumption is 0.8 seconds on average, but occasionally there are abnormal points of 1.5 seconds. Extract the fluctuation characteristics of these data (standard deviation of wakeup response time 0.08 seconds, temperature sensitivity of storage occupancy rate 0.5% / °C), and mark the time sequence of abnormal point occurrence (such as 3 times of 1.5 seconds time consumption all occur when humidity > 70%).

[0207] Then, compare the data with the S1 preset reference: the wakeup response time reference is ≤2.5 seconds (currently meets), but the fluctuation characteristics show that it can be compressed; the storage occupancy rate reference is ≤30% (currently safe), but the temperature sensitivity needs to be optimized; the abnormal point (1.5 seconds) exceeds the reference 0.8±0.2 seconds, and all coefficients are associated with the corresponding timestamp. Calculate the combined optimization coefficient and abnormal correction coefficient, the specific calculation method is as follows:

[0208] Combined optimization coefficient calculation method:

[0209] Hibernation duration optimization coefficient = (target hibernation duration - current average hibernation duration) / current average hibernation duration; wherein the target hibernation duration is determined based on "task planning interval - reserved start-up time" (the reserved start-up time is a fixed value of 5 minutes);

[0210] Data compression ratio optimization coefficient = (target compression ratio - current compression ratio) / current compression ratio; wherein the target compression ratio is dynamically adjusted based on "storage occupancy rate reference - current storage occupancy rate" (the storage occupancy rate reference is preset to 30%);

[0211] In the example, the task planning interval is 30 minutes during the day, the target hibernation duration = 30 minutes - 5 minutes = 25 minutes, the current average hibernation duration is 28 minutes, so the daytime hibernation duration optimization coefficient = (25-28) / 28 ≈ -0.107; the current compression ratio of the environment log is 1:1, the current storage occupancy rate is 25%, the target compression ratio = 1: (1+ (30%-25%)) = 1:1.05, and the data compression ratio optimization coefficient = (1.05-1) / 1 = +0.05.

[0212] Abnormal correction coefficient calculation method:

[0213] Abnormal correction coefficient = (actual time consumption in abnormal scenario - benchmark time consumption) / benchmark time consumption x (-1); wherein, the benchmark time consumption is the average time consumption of the step in normal scenario, and the abnormal scenario is defined by the "feature threshold" in historical data (e.g., humidity > 70%, 4G signal strength <-85dBm is determined as an abnormal scenario);

[0214] In an example, the normal scenario extension layer data reading benchmark time consumption is 0.8 seconds, and the abnormal time consumption is 1.5 seconds when the humidity is > 70%, so the abnormal correction coefficient = (1.5-0.8) / 0.8 x (-1) = -0.875.

[0215] Supplementary benchmark parameter determination method: all benchmark parameters (such as standard sleep duration, upper limit of wake-up time consumption, storage occupancy rate benchmark, and average time consumption in normal scenario) are determined by the following method: 1) preset initial value based on embedded system hardware performance parameters (such as main processing unit operation speed, non-volatile storage unit read / write speed); 2) the final benchmark parameters are determined and stored in the non-volatile storage unit by the historical data mean value calibration of 3 pre-running periods, and dynamic adjustment is supported by subsequent update set.

[0216] Supplementary feature and coefficient mapping logic: the mapping relationship between the extracted historical data fluctuation features (such as wake-up interval fluctuation, temperature sensitivity) and the optimization coefficient is determined by the "feature threshold - coefficient gradient" table, for example: when the wake-up interval fluctuation is ≤±2 minutes, the optimization coefficient gradient is ±0.05; when the fluctuation is > ±2 minutes and ≤±5 minutes, the coefficient gradient is ±0.10, and the gradient value is preset and can be calibrated by pre-running data.

[0217] Subsequently, a time effectiveness label is added to the optimization coefficient ("VALID_7D" represents 7 days effective), and an abnormal correction coefficient is added to the type identification ("HUMIDITY_RELATED"), and a hierarchical update set is generated: the core layer update set contains the wake-up advance amount adjustment parameter, the extension layer update set contains the log compression ratio and pre-reading mechanism parameter, and both are marked with a time sequence ("CYCLE_30" represents data based on 30 cycles). Finally, the main processing unit transmits the update set to the auxiliary processing unit, and the auxiliary processing unit updates the wake-up threshold parameter (advance amount 80ms) at address 0x0001 and the log configuration at address 0x0200 in EEPROM according to the "core layer priority" principle, adds a note to the parameter item corresponding to the abnormal type identification, and feeds back "core layer update completed, extension layer compression ratio effective", to ensure that the subsequent sleep-wake cycle is more adaptive to the on-site environment, and forms a closed loop with the overall process. Through hierarchical analysis of historical data, accurate optimization of parameters is achieved, time effectiveness label and abnormal identification ensure the effectiveness of the update, main and auxiliary collaborative update improves the self-adaptive ability of the system, and enhances the long-term running stability.

[0218] Reference Figure 2The application relates to an embedded system low-power control device, which is applied to an embedded system low-power control equipment, and the equipment comprises a main processing unit, an auxiliary processing unit and a low-power monitoring module; the main processing unit and the auxiliary processing unit are connected through a communication interface and are electrically connected with a nonvolatile storage unit; and the auxiliary processing unit is electrically connected with the low-power monitoring module; the device comprises:

[0219] A parameter transmission storage module 100 comprises a parameter generation sub-module on the main processing unit side and a parameter check storage sub-module on the auxiliary processing unit side; the parameter generation sub-module is used for generating operation parameters containing check information and transmitting the operation parameters to the auxiliary processing unit through a communication interface; the parameter check storage sub-module is used for storing the operation parameters to a nonvolatile storage unit after check; the auxiliary processing unit initializes a low-power monitoring module according to the operation parameters and feeds back a completion signal to a main system; wherein the main system is a core operation system in an embedded system which is dominated by the main processing unit;

[0220] A state monitoring and judging module 200 comprises a state sending sub-module on the main processing unit side and a condition judging sub-module on the auxiliary processing unit side; the state sending sub-module is used for sending operation state data containing task states, environment data and time limit information to the auxiliary processing unit; the condition judging sub-module is used for judging dormancy conditions by combining a wake-up threshold in the stored operation parameters and the operation state data, generating corresponding instructions and transmitting the instructions to the main processing unit;

[0221] A dormancy control module 300 comprises a data storage sub-module on the main processing unit side and a power supply control sub-module on the auxiliary processing unit side; the data storage sub-module is used for responding to a dormancy instruction, writing key data before dormancy into a nonvolatile storage unit in layers and feeding back data identification and a log; the power supply control sub-module is used for checking data identification matching and log integrity, controlling a power supply loop of the main processing unit to be cut off after check, and making the auxiliary processing unit enter a low-power state after binding related data, wherein the key data is necessary data for restoring the system to a state before dormancy, including task execution progress, core configuration parameters, data storage address index and operation state data directly related to the state before dormancy; the related data is associated data of the data identification, the log abstract and the state parameters of the main processing unit before dormancy, and the state parameters of the main processing unit include power information, task completion marks and operation mode parameters;

[0222] The wake-up recovery module 400 comprises an event monitoring sub-module on the auxiliary processing unit side and a state recovery sub-module on the main processing unit side; the event monitoring sub-module is configured to monitor a wake-up event and verify that the wake-up event is passed, control the main processing unit power supply loop to be turned on, and transmit wake-up information and a data storage location; and the state recovery sub-module is configured to read key data before hibernation and restore the key data to a state before hibernation.

[0223] The parameter update feedback module 500 comprises a coefficient calculation sub-module on the main processing unit side and a parameter update sub-module on the auxiliary processing unit side; the coefficient calculation sub-module is configured to extract features based on historical data of multiple hibernation-wake-up cycles, calculate optimization coefficients, generate an update set, and synchronize the update set to the auxiliary processing unit; and the parameter update sub-module is configured to update stored operating parameters and feed back to the main processing unit.

[0224] Further, the parameter transmission and storage module 100 comprises:

[0225] The operating parameter generation and transmission unit is configured to generate operating parameters containing check information, package the operating parameters in a preset format by the main processing unit, attach a storage address identifier, and then transmit the operating parameters to the auxiliary processing unit through a communication interface.

[0226] The operating parameter receiving and storage unit is configured to receive and reorganize the operating parameters, verify the check information by the auxiliary processing unit, and store the operating parameters in a non-volatile storage unit according to the address identifier.

[0227] The operating parameter reading and comparison unit is configured to read the parameters from the corresponding address in the non-volatile storage unit, package the parameters in the original format by the auxiliary processing unit, recalculate the check information, and compare the packaged results with the original transmission parameters and the two check information.

[0228] The parameter verification and feedback unit is configured to initialize the low-power monitoring module according to the operating parameters and feed back a completion signal to the main processing unit if the comparison is consistent, or request retransmission if the comparison is inconsistent, and the operation is performed by the auxiliary processing unit.

[0229] Further, the operating parameter reading and comparison unit comprises:

[0230] The parameter reading and feature generation sub-unit is configured to read the parameters from the corresponding address in the non-volatile storage unit, package the parameters in the original format, recalculate the check information, and extract features from the packaged parameters to generate a parameter feature value.

[0231] The parameter encryption and association sub-unit is configured to encrypt the packaged parameters, the recalculated check information, and the parameter feature value, divide the packaged parameters into main parameters and auxiliary parameters, and associate the main parameters and the auxiliary parameters with the check information, respectively.

[0232] The encrypted storage and priority verification subunit is used to store encrypted content and prioritize the verification of the matching between the master parameters and the verification information.

[0233] The parameter improvement and consistency confirmation subunit is used to decrypt and call auxiliary parameters to improve the main parameters after the main parameters have been verified, and to confirm the parameter consistency by combining the parameter characteristic values.

[0234] Furthermore, the aforementioned state monitoring and determination module 200 includes:

[0235] The operation status data sending unit is used by the main processing unit to send operation status data, including task status, environmental data, credibility identifier, timeliness information and task planning, to the auxiliary processing unit.

[0236] The hibernation condition determination unit is used by the auxiliary processing unit to verify the credibility identifier in the running status data, and to comprehensively determine whether the preset hibernation conditions are met by combining the wake-up threshold, effective time information and task planning in the stored running parameters.

[0237] The judgment instruction generation unit is used by the auxiliary processing unit to generate instructions containing data storage identifiers, credibility and timeliness verification results, and task planning association descriptions based on the judgment results.

[0238] The instruction execution and result feedback unit is used by the main processing unit to receive instructions, confirm the validity of the data and the matching of the task plan based on the verification results, execute the instructions, and feed back the results to the auxiliary processing unit.

[0239] Furthermore, the aforementioned hibernation condition determination unit includes:

[0240] The trusted data filtering and preprocessing subunit is used to verify the trustworthiness identifiers in the running status data, filter trusted data, extract key parameters and timestamps, check data integrity and fill in missing items.

[0241] The parameter comparison and timeliness analysis subunit is used to compare key parameters with the wake-up threshold in the stored running parameters, mark abnormal items, remove expired abnormal items by combining valid timeliness information, and calculate the remaining valid time of the data.

[0242] The hibernation feasibility pre-assessment subunit is used to compare the remaining effective time with the shortest hibernation time, taking into account the next start time and resource requirements of the task plan.

[0243] The anomaly and resource impact judgment subunit is used to determine the impact of anomalies and whether resources support hibernation.

[0244] The hibernation condition comprehensive judgment subunit is used to comprehensively determine whether the hibernation conditions are met by combining the above results, and generate a result containing the judgment criteria.

[0245] Furthermore, the aforementioned determination instruction generation unit includes:

[0246] The data storage mark generation subunit is configured to generate a data storage mark containing a storage path, a data type, and a storage priority according to the determination result, and to distinguish between mandatory data and optional data, wherein whether the data is necessary for the system to be restored to the state before hibernation after hibernation is taken as a distinguishing standard; the mandatory data is indispensable for the system to be restored to the state, including task execution progress, core configuration parameters, data storage address index, and key operating parameters of the main processing unit; and the optional data is auxiliary analysis or non-essential supplementary data, including environmental data details and non-core task logs.

[0247] The verification result processing subunit is configured to generate a verification summary, a state code, and a report segment based on the credibility and timeliness verification result.

[0248] The task planning association subunit is configured to extract key nodes in combination with task planning association instructions, generate an association index, and mark a time node.

[0249] The instruction integration generation subunit is configured to integrate the data storage mark, the verification result, and the key node instruction, and to form an instruction by adding integrated verification information.

[0250] Further, the task planning association subunit includes:

[0251] The key node identification subunit is configured to parse the task planning association instructions, and to identify a task start node, a task end node, a main task node, a subtask node, a resource allocation node, and a state conversion node as key nodes.

[0252] The association index generation subunit is configured to assign a unique identifier to the key nodes, generate a hierarchical association index to mark a subordinate relationship, and generate a state association index to record conversion logic.

[0253] The time node extraction and marking subunit is configured to extract and mark an execution time, a planned execution point, and an expected state conversion time of each node as time nodes.

[0254] The association table arrangement subunit is configured to arrange the key nodes, the hierarchical association index, the state association index, and the time nodes into an association table according to a corresponding relationship.

[0255] The association table arrangement subunit is configured to arrange the association table according to a subordinate relationship and state conversion logic.

[0256] Further, the hibernation control module 300 includes:

[0257] The key data hierarchical storage unit is configured to write the key data before hibernation into the non-volatile storage unit by the main processing unit, to generate a data identifier and a storage log of each layer, and to feed back the data identifier, a log summary, and a hierarchical storage result to the auxiliary processing unit.

[0258] The storage verification and power supply control unit is configured to verify the data identifier matching and the log digest integrity by the auxiliary processing unit, and to cut off the power supply loop of the main processing unit after confirming the completion of the storage of each layer of data.

[0259] The data and state parameter binding unit is configured to bind the data identifier, the log digest and the state parameter of the main processing unit according to the hierarchical correspondence of the auxiliary processing unit.

[0260] The associated record integration and low-power consumption entering unit is configured to integrate the associated data hierarchical information and the binding record, and to enter the low-power consumption state by carrying the associated record.

[0261] Further, the storage verification and power supply control unit comprises:

[0262] The hierarchical data receiving and sorting subunit is configured to receive the data identifier, the log digest and the hierarchical storage result, to distinguish the core layer and the extended layer, to sort and allocate the verification serial number according to the layer level.

[0263] The hierarchical data verification subunit is configured to verify the data identifier matching and the log digest format compliance of the core layer in sequence priority, to verify the extended layer, and to record the verification result.

[0264] The power supply cut-off permission generation subunit is configured to generate the power supply cut-off permission signal after confirming that all layers pass the verification and the digest is complete.

[0265] The power supply loop control subunit is configured to cut off the power supply loop of the main processing unit according to the permission signal.

[0266] Further, the hierarchical data receiving and sorting subunit comprises:

[0267] The layer level type distinguishing subunit is configured to receive the data identifier, the log digest and the hierarchical storage result, to identify the layer type identifier and the irreplaceable identifier, to extract the core layer containing the irreplaceable identifier and the corresponding type, and to take the rest as the extended layer.

[0268] The layer level sorting subunit is configured to extract the storage time stamp and the capacity information of the core layer and the extended layer, and to sort according to the time stamp and the capacity from large to small.

[0269] The verification serial number allocation and sequence forming subunit is configured to allocate the verification serial number containing the time stamp segment and the capacity proportion for each layer after sorting, to associate the layer type identifier, the sorting result and the verification serial number to form the hierarchical verification sequence.

[0270] Further, the data and state parameter binding unit comprises:

[0271] The hierarchical identification identification and information extraction subunit is used for identifying data hierarchical identification, distinguishing core layer and extension layer, extracting each layer data identification feature information and log abstract key information;

[0272] The hierarchical verification and parameter binding subunit is used for generating a combined verification value for the core layer and the extension layer respectively, binding the feature information, the key information, the verification value and the corresponding layer main processing unit state parameter, and adding a hierarchical type label;

[0273] The correlation key generation and record establishment subunit is used for adding the hierarchical identification as a correlation key for each binding group, and establishing the association record of the type label and the binding group;

[0274] The association record storage subunit is used for storing the association record and all binding groups in order of hierarchical identification.

[0275] Further, the above-mentioned correlation key generation and record establishment subunit comprises:

[0276] The correlation key generation embedded subunit is used for generating a unique hierarchical identification containing hierarchical information and a verification bit for each binding group as a correlation key embedded in the binding group;

[0277] The label feature extraction subunit is used for identifying the attribute features of the type label and classifying, and extracting the features of the classified label;

[0278] The label and binding group association subunit is used for associating the type label and the corresponding binding group by comparing the correlation key features and the label features, and generating the association record;

[0279] The association record verification storage subunit is used for adding a key verification code to the association record, and storing the corresponding table of the label features and the association record.

[0280] Further, the above-mentioned label feature extraction subunit comprises:

[0281] The label attribute feature extraction subunit is used for scanning the type label, extracting the basic attribute, dynamic and static attribute and association attribute feature, and determining the association object and the association strength;

[0282] The two-dimensional classification matrix forming subunit is used for classifying according to the association strength, combining the dynamic and static attribute division, and forming an association-dynamic and static two-dimensional classification matrix;

[0283] The same label feature extraction subunit is used for extracting the common and differentiated features and the association feature coefficient for the same label in the classification matrix;

[0284] The feature data correlation binding subunit is used for associating the hierarchical classification result and the feature extraction data, and binding the two-dimensional classification and the feature coefficient.

[0285] Further, the above-mentioned association record integration and low-power entry unit comprises:

[0286] a hierarchy and binding information extraction subunit for extracting hierarchy identifiers, timestamps and priorities in the data hierarchy information, and extracting association keys, time markers and priority tags in the binding records;

[0287] a multi-dimensional association generation subunit for generating association indexes according to the matching of the hierarchy identifiers and the association keys, generating a time sequence association table according to the association of the timestamps and the time markers, and generating an association queue according to the correspondence of the priorities and the priority tags;

[0288] a verification and validity identification adding subunit for adding integrity verification identifications and time validity identifications to the association indexes, the time sequence association table and the association queue;

[0289] a low-power entry subunit for carrying the association indexes, the time sequence association table and the association queue with the integrity verification identifications and the time validity identifications into a low-power state.

[0290] Further, the above-mentioned wake-up recovery module 400 comprises:

[0291] a wake-up event monitoring and information extraction unit for monitoring wake-up event related signals and data, starting a wake-up process when a trigger condition is met, recording a trigger time and extracting a trigger feature;

[0292] a double verification and power supply control unit for comparing the trigger time with a wake-up permission time period, matching the trigger feature with a wake-up feature library, and turning on a main processing unit power supply loop after double verification;

[0293] a wake-up information transmission unit for transmitting the trigger time, the feature matching result and the storage path of the key data before hibernation to the main processing unit;

[0294] a data reading and state recovery unit for reading corresponding hierarchical data and state parameters according to the storage path, and recovering to a state before hibernation according to the parameters.

[0295] Further, the above-mentioned double verification and power supply control unit comprises:

[0296] a trigger information processing and hierarchy division subunit for extracting the trigger time and the trigger feature, dividing the wake-up permission time period and the feature library according to hierarchy, adding a timestamp to the trigger time, and generating a feature code from the trigger feature;

[0297] a hierarchy comparison and confidence generation subunit for comparing the timestamp with the corresponding hierarchy permission time period, comparing the feature code with the corresponding hierarchy feature library, and generating a confidence value of the two;

[0298] The level consistency and confidence verification subunit is used for associating a time period level and a feature level, verifying whether consistency and confidence meet standards;

[0299] The power-on control subunit is used for generating a power-on signal to turn on a power supply loop of the main processing unit when the level is consistent and the confidence meets the standard.

[0300] Further, the trigger information processing and level division subunit comprises:

[0301] The trigger information extraction and priority marking subunit is used for identifying a trigger source type and an occurrence sequence, extracting a corresponding trigger time and feature details, and marking a trigger priority in combination with an environmental parameter;

[0302] The permission time period and feature library level division subunit is used for dividing a wake-up permission time period into a core layer and an extended layer according to an importance level, and dividing a feature library according to a type matching degree level, and synchronously associating a sequence and a priority;

[0303] The timestamp and feature code generation subunit is used for adding an environmental timestamp containing a level identifier, a type and a sequence to a trigger time, and generating a matching degree feature code for a trigger feature;

[0304] The level and parameter binding subunit is used for binding a timestamp level and a feature code level, associating an environmental parameter and a priority, and forming a level matching set.

[0305] Further, the wake-up information transmission unit comprises:

[0306] The information integration and association subunit is used for integrating a trigger time, a feature matching result and a storage path of key data before hibernation, binding the trigger time and the storage path, adding a credibility label to the feature matching result and associating a path index;

[0307] The composite information generation subunit is used for extracting a trigger time slice, a feature matching result feature slice, and splicing the storage path into composite information, adding a splicing identifier and a transmission check code;

[0308] The mapping set generation subunit is used for generating a mapping set of a binding relationship, an association index and a splicing identifier;

[0309] The information transmission subunit is used for transmitting the composite information containing the transmission check code and the mapping set to the main processing unit, so that the main processing unit verifies the check code and analyzes the information according to the mapping set.

[0310] Further, the composite information generation subunit comprises:

[0311] The information slice interception subunit is used for intercepting a key timestamp slice of a trigger time, a feature hash slice of a feature matching result, and extracting a permission and a version slice of a storage path.

[0312] The composite information splicing subunit is configured to splice the preset separator to generate the composite information in the order of timestamp-feature hash-permission-version-path.

[0313] The splicing identifier adding subunit is configured to add the splicing identifier including the splicing order and the separator type to the composite information.

[0314] The check code calculating and adding subunit is configured to calculate the overall transmission check code based on the content of each segment, and add the check code to the composite information.

[0315] Further, the parameter updating feedback module 500 includes:

[0316] The historical data processing and feature extraction unit is configured to distinguish the core and extended historical data, perform statistics based on the historical data of multiple sleep-wake cycles, extract the data fluctuation features and abnormal points, and record the time sequence.

[0317] The parameter comparison and coefficient calculation unit is configured to compare the core and extended data, the fluctuation features and the abnormal points with corresponding references respectively, calculate the combination optimization coefficient and the abnormal correction coefficient, and associate the time stamp.

[0318] The hierarchical update set generation unit is configured to add the time effectiveness label and the abnormal type identifier to the optimization coefficient, and generate the hierarchical update set with the time sequence label.

[0319] The update transmission and result feedback unit is configured to transmit the update set to the auxiliary processing unit, and the auxiliary processing unit updates the stored operation parameters according to the identifier classification, and feeds back the update result and the abnormal processing condition to the main processing unit.

[0320] Reference Figure 3 In the embodiments of the present application, a computer device, which can be a server, is also provided, and the internal structure of the computer device can be as shown in Figure 3The computer device includes a processor, a memory, a network interface and a database connected by a bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores operations, computer programs and databases. The internal memory provides an environment for the operation and computer program in the non-volatile storage medium. The database of the computer device is used to store data such as the embedded system low-power control method. The network interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement an embedded system low-power control method, which is applied to an embedded system low-power control device including a main processing unit, an auxiliary processing unit and a low-power monitoring module. The main processing unit and the auxiliary processing unit are connected through a communication interface and are electrically connected to a non-volatile storage unit. The auxiliary processing unit is electrically connected to the low-power monitoring module. The method includes: the main processing unit generates running parameters containing check information, transmits the running parameters to the auxiliary processing unit through the communication interface, stores the running parameters in the non-volatile storage unit after the running parameters are verified by the auxiliary processing unit, and the auxiliary processing unit initializes the low-power monitoring module according to the running parameters and feeds back a completion signal to the main system, wherein the main system is a core operation system in the embedded system dominated by the main processing unit; the main processing unit sends running state data containing task state, environment data and time limit information to the auxiliary processing unit, the auxiliary processing unit determines the sleep condition by combining the wake-up threshold in the stored running parameters and the running state data, generates corresponding instructions and transmits the instructions to the main processing unit; the main processing unit responds to the sleep instructions, layer-by-layer writes key data before sleep to the non-volatile storage unit, and feeds back data identification and logs, which are verified by the auxiliary processing unit, and the auxiliary processing unit controls the power supply loop of the main processing unit to be cut off, and the auxiliary processing unit enters a low-power state after binding related data; the key data is necessary data for restoring the system to the state before sleep, including task execution progress, core configuration parameters, data storage address index, and is directly related to the running state data before sleep; the related data is associated data of the data identification, log summary and state parameters of the main processing unit before sleep, and the state parameters of the main processing unit include power information, task completion mark and running mode parameters; after the auxiliary processing unit monitors the wake-up event and verifies it, the auxiliary processing unit controls the power supply loop of the main processing unit to be turned on, transmits wake-up information and data storage position to the main processing unit, and the main processing unit reads the key data before sleep to restore to the state before sleep; the main processing unit extracts features based on historical data of multiple sleep-wake cycles, calculates an optimization coefficient to generate an updated set and synchronizes the updated set to the auxiliary processing unit, and the auxiliary processing unit updates the stored running parameters and feeds back to the main processing unit.

[0321] An embodiment of the present application also provides a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement an embedded system low-power consumption control method. The embedded system low-power consumption control method is applied to an embedded system low-power consumption control device. The device includes a main processing unit, an auxiliary processing unit, and a low-power consumption monitoring module. The main processing unit and the auxiliary processing unit are connected through a communication interface and are electrically connected to a non-volatile storage unit. The auxiliary processing unit is electrically connected to the low-power consumption monitoring module. The method includes the following steps. The main processing unit generates running parameters containing check information, and transmits the running parameters to the auxiliary processing unit through the communication interface. The running parameters are stored in the non-volatile storage unit after being checked by the auxiliary processing unit. The auxiliary processing unit initializes the low-power consumption monitoring module according to the running parameters and feeds back a completion signal to the main system. The main system is a core operation system in the embedded system which is dominated by the main processing unit. The main processing unit sends running state data containing task state, environment data, and time limit information to the auxiliary processing unit. The auxiliary processing unit determines a sleep condition according to the wake-up threshold in the stored running parameters and the running state data, generates corresponding instructions, and transmits the instructions to the main processing unit. The main processing unit writes key data before sleep into the non-volatile storage unit in layers in response to the sleep instructions, and feeds back data identifiers and logs. After being checked by the auxiliary processing unit, the power supply loop of the main processing unit is cut off by the auxiliary processing unit. The auxiliary processing unit enters a low-power consumption state after binding and recording related data. The key data is necessary data for restoring the system to the state before sleep, including task execution progress, core configuration parameters, and data storage address index, and is directly related to the running state data before sleep. The related data is associated data of the data identifiers, log abstracts, and state parameters of the main processing unit before sleep. The state parameters of the main processing unit include power information, task completion markers, and running mode parameters. After detecting a wake-up event and verifying it, the auxiliary processing unit controls the power supply loop of the main processing unit to be conductive, transmits wake-up information and data storage positions to the main processing unit, and reads the key data before sleep to restore the system to the state before sleep. The main processing unit extracts features based on historical data of multiple sleep-wake cycles, calculates an optimization coefficient to generate an updated set, and synchronizes the updated set to the auxiliary processing unit. The auxiliary processing unit updates the stored running parameters and feeds back to the main processing unit.

[0322] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiment methods can be included. Any reference to memory, storage, database or other medium provided by the present application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0323] The above only describes the preferred embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation using the content of the present application specification and drawings, or direct or indirect application in other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A low-power control method for an embedded system, characterized in that, The embedded system low-power control method is applied to an embedded system low-power control device, the device including a main processing unit, an auxiliary processing unit, and a low-power monitoring module. The main processing unit and the auxiliary processing unit are connected via a communication interface and are both electrically connected to a non-volatile memory unit. The auxiliary processing unit is electrically connected to the low-power monitoring module. The method includes: The main processing unit generates operating parameters containing verification information and transmits them to the auxiliary processing unit through the communication interface. After the auxiliary processing unit verifies the parameters, the parameters are stored in the non-volatile storage unit. The auxiliary processing unit initializes the low-power monitoring module based on the operating parameters and sends a completion signal back to the main system. The main system is the core computing system of the embedded system dominated by the main processing unit. The main processing unit sends operation status data containing task status, environmental data, and timeliness information to the auxiliary processing unit. The auxiliary processing unit combines the wake-up threshold in the stored operation parameters and the operation status data to determine the hibernation conditions, and generates corresponding instructions to transmit to the main processing unit. The main processing unit responds to the hibernation command, writes the key data before hibernation into the non-volatile storage unit in layers, and feeds back the data identifier and log. After the auxiliary processing unit verifies and passes the verification, the auxiliary processing unit controls the power supply circuit of the main processing unit to be cut off. The auxiliary processing unit binds and records the relevant data and then enters a low-power state. Among them, the key data is the data necessary for the system to be restored to the state before hibernation after hibernation; the relevant data is the data associated with the state parameters of the main processing unit before hibernation. After the auxiliary processing unit detects and verifies the wake-up event, it controls the main processing unit to turn on the power supply circuit and transmits the wake-up information and data storage location to the main processing unit. The main processing unit reads the key data before hibernation and restores it to the state before hibernation. The main processing unit extracts features based on historical data from multiple sleep-wake cycles, calculates optimization coefficients, generates an update set, and synchronizes it to the auxiliary processing unit. The auxiliary processing unit updates the stored operating parameters and feeds them back to the main processing unit.

2. The low-power control method for embedded systems according to claim 1, characterized in that, The main processing unit sends operation status data containing task status, environmental data, and timeliness information to the auxiliary processing unit. The auxiliary processing unit, in conjunction with the wake-up threshold in the stored operation parameters and the operation status data, determines the sleep conditions and generates corresponding instructions to transmit to the main processing unit. The steps include: The main processing unit sends runtime status data, including task status, environmental data, credibility identifier, timeliness information, and task planning, to the auxiliary processing unit. The auxiliary processing unit verifies the credibility identifier in the running status data and, in combination with the wake-up threshold, effective timeliness information and task planning in the stored running parameters, comprehensively determines whether the preset hibernation conditions are met. The auxiliary processing unit generates instructions based on the judgment results, including data storage identifiers, credibility and timeliness verification results, and task planning association descriptions. The main processing unit receives the instruction, confirms the validity of the data and its matching with the task plan based on the verification results, executes the instruction, and feeds back the results to the auxiliary processing unit.

3. The low-power control method for embedded systems according to claim 2, characterized in that, The auxiliary processing unit generates instructions containing data storage identifiers, credibility and timeliness verification results, and task planning association descriptions based on the judgment results, including: Based on the judgment results, a data storage identifier is generated, which includes the storage path, data type, and storage priority, to distinguish between mandatory and optional data. The distinction is based on whether the data is necessary for the system to recover to its pre-hibernation state after hibernation. Mandatory data is data that is indispensable for restoring the system state. Optional data is supplementary data that is used for auxiliary analysis or is not necessary. Generate a verification summary, status code, and report fragment from the credibility and timeliness verification results; Extract key nodes by combining task planning and related descriptions, and generate related indexes and labeled time nodes; Integrate data storage identifiers, verification results, and key node descriptions, and attach integrated verification information to form instructions.

4. The low-power control method for embedded systems according to claim 3, characterized in that, The steps of extracting key nodes by combining task planning and association descriptions, generating association indexes, and annotating time nodes include: Analyze the task planning association description and identify the task start, end, main task, sub-task, resource allocation and state transition nodes as key nodes; Assign unique identifiers to key nodes, generate hierarchical association indexes to mark subordinate relationships, and generate state association index records for conversion logic; Extract the execution time, planned execution point, and expected state transition time of each node as time nodes and label them; Organize key nodes, hierarchical association indexes, status association indexes, and time nodes into an association table according to their corresponding relationships; The association table is arranged according to the hierarchy and state transition logic.

5. The low-power control method for embedded systems according to claim 1, characterized in that, The main processing unit responds to the hibernation command, writes the key data before hibernation into the non-volatile storage unit in layers, and feeds back the data identifier and log. After the auxiliary processing unit verifies the data, the auxiliary processing unit controls the power supply circuit of the main processing unit to be cut off. The auxiliary processing unit binds and records the relevant data and then enters a low-power state. The steps include: The main processing unit writes the key data before hibernation into the non-volatile storage unit in layers, generates data identifiers and storage logs for each layer, and feeds back the data identifiers, log summaries and layered storage results to the auxiliary processing unit. The auxiliary processing unit verifies the matching of data identifiers and the integrity of log summaries. After confirming that the data storage at each layer is completed, the auxiliary processing unit cuts off the power supply circuit to the main processing unit. The auxiliary processing unit binds and records data identifiers, log summaries, and main processing unit status parameters according to the data hierarchical correspondence. The auxiliary processing unit associates hierarchical data information with bound records and generates associated records. The auxiliary processing unit carries the associated record into a low-power state.

6. The low-power control method for embedded systems according to claim 5, characterized in that, The auxiliary processing unit binds and records data identifiers, log summaries, and main processing unit status parameters according to the data hierarchical correspondence, including: Identify data layer identifiers, distinguish between the core layer and the extended layer, and extract data identifier feature information and key information from log summaries for each layer; Generate merged verification values ​​for the core layer and the extension layer respectively, bind feature information, key information, verification values ​​and corresponding layer main processing unit status parameters, and add layer type labels; Add a hierarchical identifier as an association key to each binding group, and establish an association record between the type label and the binding group; Related records and all binding groups are stored in hierarchical order.

7. The low-power control method for embedded systems according to claim 6, characterized in that, The step of adding a hierarchical identifier as an association key to each binding group and establishing an association record between the type label and the binding group includes: A unique hierarchical identifier containing hierarchical information and a check bit is generated for each binding group and embedded as an association key into the binding group; Identify and classify the attribute features of type labels, and extract features from the classified labels; By comparing the associated key features with the tag features, the associated type tag is associated with the corresponding binding group, and an associated record is generated; Add a key verification code to the associated records and store a table that maps tag features to associated records.

8. The low-power control method for embedded systems according to claim 7, characterized in that, The steps of identifying and classifying the attribute features of type labels, and extracting features from the classified labels, include: Scan type labels to extract basic attributes, dynamic and static attributes, and related attribute features to determine related objects and the strength of the association. Based on the strength of association, and combined with dynamic and static attributes, a two-dimensional classification matrix of association and dynamic / static attributes is formed. For labels of the same type within the classification matrix, extract common and differential features and correlation coefficients; The hierarchical classification results are correlated with the feature extraction data, and the two-dimensional classification and feature coefficients are bound together.

9. The low-power control method for embedded systems according to claim 5, characterized in that, The auxiliary processing unit verifies the matching of data identifiers and the integrity of log digests. After confirming that data storage at each layer is complete, the auxiliary processing unit disconnects the power supply circuit to the main processing unit, including: Receive data identifiers, log summaries, and hierarchical storage results; identify the type identifiers and non-replaceable identifiers of each layer; extract the core layer containing non-replaceable identifiers and their corresponding types; and use the rest as extension layers. Extract the storage timestamps and capacity information of the core layer and the extension layer, and sort them by timestamp order and capacity from largest to smallest. Each sorted layer is assigned a verification sequence number containing timestamp segments and capacity percentages. The layer type identifier, sorting results, and verification sequence numbers are linked to form a layered verification sequence.

10. A low-power control device for an embedded system, characterized in that, The embedded system low-power control device is applied to an embedded system low-power control equipment. The equipment includes a main processing unit, an auxiliary processing unit, and a low-power monitoring module. The main processing unit and the auxiliary processing unit are connected via a communication interface and are both electrically connected to a non-volatile memory unit. The auxiliary processing unit is electrically connected to the low-power monitoring module. The device includes: The parameter transmission and storage module includes a parameter generation submodule on the main processing unit side and a parameter verification and storage submodule on the auxiliary processing unit side. The parameter generation submodule generates operating parameters containing verification information and transmits them to the auxiliary processing unit through a communication interface. The parameter verification and storage submodule receives the operating parameters, verifies them, and stores them in a non-volatile storage unit. The auxiliary processing unit initializes a low-power monitoring module based on the operating parameters and sends a completion signal back to the main system. The main system is the core computing system of the embedded system, which is dominated by the main processing unit. The status monitoring and determination module includes a status sending submodule on the main processing unit side and a condition determination submodule on the auxiliary processing unit side. The status sending submodule is used to send running status data containing task status, environmental data and timeliness information to the auxiliary processing unit. The condition determination submodule is used to determine the sleep condition by combining the wake-up threshold in the stored running parameters and the running status data, and generate a corresponding instruction to be transmitted to the main processing unit. The hibernation control module includes a data storage submodule on the main processing unit side and a power supply control submodule on the auxiliary processing unit side. The data storage submodule responds to hibernation commands by writing critical data prior to hibernation into a non-volatile storage unit and feeding back data identifiers and logs. The power supply control submodule verifies the matching of data identifiers and the integrity of logs. Upon successful verification, it cuts off the power supply circuit to the main processing unit, and the auxiliary processing unit binds and records the relevant data before entering a low-power state. The critical data is necessary for restoring the system to its pre-hibernation state after hibernation; the relevant data is data associated with the pre-hibernation state parameters of the main processing unit. The wake-up recovery module includes an event monitoring submodule on the auxiliary processing unit side and a state recovery submodule on the main processing unit side. The event monitoring submodule is used to monitor wake-up events and, after verification, control the main processing unit power supply circuit to be turned on and transmit wake-up information and data storage location. The state recovery submodule is used to read key data before hibernation and restore it to the state before hibernation. The parameter update feedback module includes a coefficient calculation submodule on the main processing unit side and a parameter update submodule on the auxiliary processing unit side. The coefficient calculation submodule is used to extract features based on historical data from multiple sleep-wake cycles, calculate and optimize coefficients to generate an update set and synchronize it to the auxiliary processing unit. The parameter update submodule is used to update the stored operating parameters and feed them back to the main processing unit.

Images