Iot device linkage control method and system based on multi-modal timing trigger

Abstract

The application relates to the technical field of Internet of Things devices, in particular to an Internet of Things device linkage control method and system based on multi-modal time sequence triggering, which comprises the following steps: acquiring sensing data and device state data, generating abnormal evolution chain data; identifying a risk stage and arranging a device action sequence; sending control information to make a target device execute an action; adjusting an unexecuted action according to feedback to obtain device execution state data; analyzing sensing data after device action to obtain recovery state data; and performing linkage control in combination with the recovery state and the device execution state. The application can distinguish between transient disturbance and persistent anomaly, match device action with risk degree, adjust subsequent action according to actual feedback, and sequentially release linkage after on-site stable recovery, so that mis-triggering, disordered execution and repeated start-stop of devices are reduced.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) devices, and more particularly to a method, system, and storage medium for IoT device linkage control based on multimodal timing triggering. Background Technology

[0002] In the charging areas for new energy vehicles in underground parking lots, charging equipment, environmental sensing equipment, and anomaly handling equipment need to operate in coordination. Vehicle charging anomalies may manifest as one or more anomalies in current, temperature, smoke, gas concentration, and image status appearing in a certain sequence and continuing to develop. Vehicle entry, dust generation, changes in ambient temperature, and load switching may also cause short-term data fluctuations. Therefore, it is necessary to accurately distinguish between transient disturbances and continuously developing anomalies in order to promptly and safely handle them.

[0003] Currently, existing technologies typically acquire data such as charging current, temperature, smoke concentration, gas concentration, and on-site images through charging piles, environmental sensors, and image acquisition equipment, and compare each data point with preset thresholds. When a data point exceeds the corresponding threshold, the control platform executes operations such as reducing charging power, disconnecting the charging circuit, triggering audible and visual alarms, and activating ventilation or evacuation instructions according to preset linkage rules.

[0004] It is evident that existing technologies primarily rely on single data points and fixed thresholds to determine anomalies, lacking joint analysis of the occurrence sequence, duration, and correlation of different anomaly signals. This can easily lead to misjudging short-term environmental disturbances as anomalies or delaying triggering of responses when anomalies continue to develop. Furthermore, existing linkage rules typically activate fixed equipment directly based on the anomaly level, lacking continuous assessment of equipment execution feedback, action dependencies, and recovery status. This may result in the activation of subsequent actions before the previous action is completed, or equipment repeatedly starting and stopping due to data fluctuations, causing inconsistencies between the linkage process and the actual on-site conditions. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies, such as the lack of joint analysis of the occurrence sequence, duration, and correlation of multiple abnormal signals, which easily leads to misjudgment of short-term disturbances or delayed identification of continuously developing anomalies. At the same time, existing linkage rules lack continuous judgment of device execution feedback, action dependencies, and recovery status, which easily leads to the premature initiation of subsequent actions or repeated start-stop of devices. Therefore, this invention proposes an IoT device linkage control method, system, and storage medium based on multimodal timing triggering.

[0006] To at least solve some of the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, the present invention provides an IoT device linkage control method based on multimodal timing triggering, comprising:

[0008] Acquire sensing data and device status data, perform anomaly timing analysis on the sensing data, and obtain anomaly evolution chain data;

[0009] The abnormal evolution chain data is subjected to stage identification to obtain risk stage data, which is then combined with the equipment status data to obtain equipment action sequence data.

[0010] Control information is sent based on the device action sequence data to cause the target device to perform the actions in the device action sequence data.

[0011] Obtain feedback data after the device executes the action, and adjust the actions of the unexecuted device based on the feedback data to obtain the device execution status data;

[0012] Acquire sensing data after device action, perform recovery status analysis on the sensing data, and obtain recovery status data;

[0013] Based on the recovery status data and the device execution status data, linkage status control is performed to obtain the linkage control result.

[0014] Preferably, anomaly time-series analysis is performed on the perceived data to obtain anomaly evolution chain data, including:

[0015] Perform time series validity analysis on the sensed data to obtain reliable time series data;

[0016] Anomaly identification is performed on the trusted time-series data to obtain anomaly evolution chain data.

[0017] Preferably, anomaly identification is performed on the trusted time-series data to obtain anomaly evolution chain data, including:

[0018] Anomaly extraction is performed on the trusted time-series data to obtain preliminary anomaly records;

[0019] Perform time-series correlation on the preliminary anomaly records to obtain anomaly evolution chain data.

[0020] Preferably, a temporal correlation is performed on the preliminary anomaly records to obtain anomaly evolution chain data, including:

[0021] The preliminary abnormal records are arranged to obtain an abnormal record sequence;

[0022] Perform a correlation determination on the abnormal record sequence to obtain abnormal evolution chain data.

[0023] Preferably, the abnormal evolution chain data is subjected to stage identification to obtain risk stage data, and then arranged with the equipment status data to obtain equipment action sequence data, including:

[0024] Stage identification is performed on the abnormal evolution chain data to obtain risk stage data;

[0025] The risk stage data and equipment status data are processed to perform action orchestration, resulting in equipment action sequence data.

[0026] Preferably, the risk stage data and equipment status data are subjected to action orchestration to obtain equipment action sequence data, including:

[0027] The candidate device actions are determined based on the aforementioned risk stage data;

[0028] The device status data is used to filter candidate device actions and arrange them to obtain device action sequence data.

[0029] Preferably, the unexecuted device actions are adjusted based on feedback data to obtain device execution status data, including:

[0030] The feedback data is used to determine the completion of the action and obtain the action completion result.

[0031] Adjust the unexecuted device actions based on the completion results of the actions to obtain device execution status data.

[0032] Preferably, the unexecuted device actions are adjusted based on the action completion result to obtain device execution status data, including:

[0033] Identify execution anomalies in the completion results of the actions to obtain abnormal action data;

[0034] Based on the abnormal action data, the actions of the unexecuted devices are adjusted to obtain the device execution status data.

[0035] Preferably, the unexecuted device actions are adjusted based on the abnormal action data to obtain device execution status data, including:

[0036] Block the dependent actions corresponding to the abnormal action data to obtain the action data to be adjusted;

[0037] By combining the risk stage data and equipment status data, the action data to be adjusted is obtained to obtain the equipment execution status data.

[0038] Secondly, the present invention provides an IoT device linkage control system based on multimodal timing triggering, comprising:

[0039] The data processing unit is used to acquire sensing data and device status data, perform anomaly time series analysis on the sensing data, and obtain anomaly evolution chain data.

[0040] An anomaly detection unit is used to identify the stage of the anomaly evolution chain data, obtain risk stage data, and arrange it with equipment status data to obtain equipment action sequence data.

[0041] The linkage execution unit is used to send control information based on the device action sequence data, so as to cause the target device to perform the actions in the device action sequence data; to obtain feedback data after the device execution, and to adjust the unexecuted device actions according to the feedback data to obtain device execution status data;

[0042] The recovery control unit is used to acquire sensing data after the device operates, perform recovery status analysis on the sensing data to obtain recovery status data, and perform linkage status control based on the recovery status data and the device execution status data to obtain linkage control results.

[0043] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method in the first aspect.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] I. This invention performs time-series validity analysis, abnormal change extraction, and time-series correlation on perceived data to determine whether different types of anomalies occur consecutively in a reasonable order. This helps to distinguish between instantaneous disturbances and gradually developing real anomalies, reduces erroneous linkages, and improves the timeliness of anomaly identification.

[0046] Second, by identifying the execution stage of the abnormal evolution chain data and combining it with equipment status data to screen and arrange candidate equipment actions, this invention helps to match equipment actions with the degree of abnormality and the status of on-site equipment, avoiding excessive linkage when the abnormality is minor and insufficient handling actions when the abnormality worsens.

[0047] Third, this invention determines whether the equipment action has actually been completed based on feedback data, and blocks and adjusts the unexecuted equipment action when an abnormality occurs. This helps to avoid directly identifying the successful transmission of control information as the completion of the on-site action, limits the impact of local equipment failure, and allows independent safety actions to continue to be executed.

[0048] Fourth, this invention performs recovery state analysis on the sensing data after the equipment operates, and combines the equipment execution state control to maintain and de-link the equipment. This helps to avoid erroneous judgments caused by active power outages, reduce repeated starting and stopping of the equipment near the judgment boundary, and orderly de-link the equipment after the on-site stability is restored. Attached Figure Description

[0049] Figure 1 This is a flowchart of the IoT device linkage control method of the present invention.

[0050] Figure 2 This is a comparison diagram of the effects of the present invention and the prior art. Detailed Implementation

[0051] To make the technical means, creative features, objectives, and effects of this invention easier to understand, the invention is further described below with reference to specific embodiments. However, the following embodiments are merely preferred embodiments of this invention and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the protection scope of this invention.

[0052] A specific embodiment of the present invention discloses an IoT device linkage control method based on multimodal timing triggering. Please refer to [link / reference]. Figure 1 This includes steps S01 to S06.

[0053] It should be noted that this embodiment is applied to the charging area of ​​new energy vehicles in an underground parking lot. The target charging location is equipped with charging piles, current detection devices, temperature sensors, smoke sensors, gas sensors, image acquisition equipment, charging circuit on / off devices, audible and visual alarms, ventilation equipment, and evacuation guidance equipment. The processing terminal is equipped with a data processing unit, an anomaly judgment unit, a linkage execution unit, and a recovery control unit, and is connected to the field equipment via an Internet of Things (IoT) communication network, ensuring that anomaly identification, equipment linkage, and status recovery all correspond to the same charging location.

[0054] Step S01: Obtain sensing data and device status data, perform anomaly timing analysis on the sensing data, and obtain anomaly evolution chain data.

[0055] Specifically, the data processing unit obtains charging voltage and charging power from the charging pile controller, charging current from the current detection device, temperature, smoke concentration and gas concentration from the temperature sensor, smoke sensor and gas sensor, and image of the target charging location from the image acquisition device to obtain sensing data. Each sensing data carries the acquisition time and charging location identifier, which helps to accurately correspond different types of sensing data to the same charging location.

[0056] The image acquisition equipment includes infrared thermal imaging equipment and visible light image acquisition equipment. Infrared thermal imaging equipment is used to acquire temperature distribution images of the target charging location, while visible light image acquisition equipment is used to acquire on-site images of the target charging location, which is helpful for identifying local high temperature phenomena and smoke phenomena respectively.

[0057] The data processing unit acquires equipment status data from charging piles, current detection devices, temperature sensors, smoke sensors, gas sensors, image acquisition equipment, charging circuit switching devices, audible and visual alarms, ventilation equipment, and evacuation guidance equipment. Equipment status data includes equipment identification, charging location identification, acquisition time, online status, fault status, and current operating status; for sensing devices, equipment status data also includes the current sampling period.

[0058] The online and fault status of sensing devices are used to determine whether the corresponding sensing data is valid. The online and fault status of execution devices are used to filter target devices that can perform the actions of candidate devices. This helps to avoid data collected by faulty devices from participating in anomaly judgment and reduces the sending of invalid control information to offline or faulty devices.

[0059] Anomaly time series analysis refers to sequentially performing time series validity analysis, anomaly change extraction, and time series correlation on sensed data. It is used to determine whether different types of anomalies occur consecutively in a reasonable order, thus avoiding reliance on whether a single data point exceeds the judgment boundary as the sole criterion. This helps distinguish between instantaneous disturbances and gradually developing true anomalies. Specifically:

[0060] S01.1 Perform time series validity analysis on the perceived data to obtain reliable time series data;

[0061] Specifically, the data processing unit divides time intervals according to a unified time axis, using the reference time of each time interval as the corresponding time. For multiple sampled values ​​generated by the same sensing device within the same time interval, the sampled value whose acquisition time is closest to the reference time is retained; if no sampled value is generated, the data of that sensing device within the corresponding time interval is marked as missing. Performing the above processing separately on different sensing devices helps to ensure that various types of sensing data retain their own meaning and are accurately mapped to a unified time position.

[0062] Subsequently, the data processing unit subtracts the data acquisition time from the time the data arrives at the processing terminal to obtain the acquisition delay, and counts the number of consecutive time intervals during which the same sensing device fails to generate a sample value, thus obtaining the consecutive missing value count. It also counts the data interruption times during normal communication conditions during device installation and debugging, and uses the maximum value among these as the maximum allowable data interruption time. Then, the maximum allowable data interruption time is divided by the sensing device's sampling period and rounded up to obtain the consecutive missing value threshold. For example, if the maximum data interruption time is 15 seconds and the sampling period is 5 seconds, then the consecutive missing value threshold is 3.

[0063] When any of the following situations occur: fault status of the corresponding sensing device's equipment status data record, consecutive missing data reaching the consecutive missing data threshold, or acquisition delay exceeding the corresponding sensing device's sampling period, the corresponding data will not be used for subsequent anomaly identification; the remaining data will be arranged according to the acquisition time to obtain reliable time-series data.

[0064] The above processing simultaneously corrects the time position between different sensing devices and eliminates faulty, severely delayed, and continuously missing data, which helps prevent invalid data from forming false abnormal sequential relationships.

[0065] S01.2 Perform anomaly identification on the trusted time series data to obtain anomaly evolution chain data; wherein, anomaly identification refers to extracting persistent abnormal changes from the trusted time series data, and then determining whether different abnormal changes belong to the same abnormal development process, thereby avoiding the direct triggering of the device by simply listing multiple independent abnormal data, which is conducive to improving the accuracy of anomaly identification.

[0066] First, perform anomaly extraction on the trusted time-series data to obtain preliminary anomaly records;

[0067] The data processing unit retrieves historical data generated during normal charging periods when the target charging location is fault-free and no abnormal handling occurs. It then removes missing and faulty data according to data type, determines the minimum and maximum values ​​of each type of historical data, and forms the corresponding normal fluctuation range. The maximum value of the normal fluctuation range is the upper limit of the normal fluctuation range, which reflects the range of data changes of the target charging location under normal operating conditions and helps to avoid identifying normal charging fluctuations as abnormalities.

[0068] The safety boundary configured in the charging area is the maximum state allowed for the corresponding equipment or location. It is determined based on at least one of the rated parameters of the charging equipment, the safety setting value of the sensing equipment, and the safety requirements of the location. The safety boundary is used to limit the maximum possible change state.

[0069] The data processing unit defines a judgment boundary between the upper limit of normal temperature fluctuation and the safety boundary. For example, the upper limit of normal temperature fluctuation is 48 degrees Celsius, and the safety boundary is 60 degrees Celsius. The midpoint of these two values, 54 degrees Celsius, is taken as the temperature judgment boundary. Once the temperature reaches 54 degrees Celsius, the judgment is further made by combining the temperature rise per unit time and the duration of the temperature rise. When all three conditions reach the corresponding thresholds, a preliminary temperature anomaly record is generated, thereby identifying anomalies before reaching the safety boundary.

[0070] For the charging current, the data processing unit reads the charging current corresponding to the current time interval from the reliable time-series data to obtain the current current value. Then, from the data generated before the anomaly judgment of the current charging task, it selects the time period in which the charging power remains stable and the charging pile does not record a fault, and calculates the arithmetic mean of the charging current during the time period to obtain the reference current.

[0071] The current variation amplitude is obtained by dividing the absolute value of the difference between the current value and the reference current by the reference current and multiplying by 100%. When the current variation amplitude continuously reaches the current variation threshold, the current duration is obtained by multiplying the corresponding number of consecutive samples by the current sampling period. When both the current variation amplitude and the current duration reach the current duration threshold, a preliminary electrical anomaly record is generated, which helps to eliminate false triggers caused by sudden changes at a single sampling point and short-term load switching.

[0072] For temperature data, the temperature rise per unit time is obtained by subtracting the temperature at the previous sampling time from the current temperature and then dividing by the time between the two samplings. The duration of continuous temperature rise is then calculated by multiplying the number of consecutive samples where the current temperature reaches the temperature judgment boundary and the temperature rise per unit time reaches the temperature rise threshold by the temperature sampling period. When the current temperature reaches the temperature judgment boundary, the temperature rise per unit time reaches the temperature rise threshold, and the duration of continuous temperature rise reaches the temperature duration threshold, a preliminary temperature anomaly record is generated. By simultaneously checking the temperature level, the temperature rise rate, and the duration, it is helpful to identify ongoing temperature anomalies and eliminate short-term temperature fluctuations.

[0073] For smoke concentration, the smoke duration is obtained by multiplying the number of samples where the smoke concentration continuously reaches the smoke judgment boundary by the smoke sampling period. When the smoke concentration reaches the smoke judgment boundary and the smoke duration reaches the smoke duration threshold, a preliminary smoke anomaly record is generated. This process only generates records when the smoke concentration is continuously abnormal, which helps to exclude single concentration increases caused by vehicle passage and short-term dust.

[0074] For gas concentration, the number of samples in which the gas concentration continuously reaches the gas judgment boundary is multiplied by the gas sampling period to obtain the gas duration. When the gas concentration reaches the gas judgment boundary and the gas duration reaches the gas duration threshold, a preliminary gas anomaly record is generated. This eliminates short-term concentration fluctuations caused by ventilation changes and helps the preliminary gas anomaly record accurately reflect persistent gas anomalies.

[0075] For image perception data, the data processing unit uses existing temperature region identification methods to process the temperature distribution images acquired by the infrared thermal imaging equipment to determine local high-temperature areas; and uses existing smoke region identification methods to process the on-site images acquired by the visible light image acquisition equipment to determine smoke regions.

[0076] Subsequently, the number of pixels in the corresponding abnormal region is divided by the total number of pixels in the target charging location region, and then multiplied by 100% to obtain the abnormal area percentage. When the abnormal area percentage continuously reaches the image area threshold, the duration of the image abnormality is obtained by multiplying the number of consecutive image frames by the image sampling period.

[0077] When the proportion of abnormal area reaches the image area threshold and the duration of image abnormality reaches the image duration threshold, a preliminary image abnormality record is generated. By using infrared temperature distribution and visible smoke features to judge image abnormalities, it is helpful to eliminate image misjudgments caused by illumination reflection, short-term occlusion and the passing of adjacent vehicles.

[0078] Among them, the current change threshold, current duration threshold, temperature rise threshold, temperature duration threshold, smoke duration threshold, gas duration threshold, image area threshold, and image duration threshold are calibrated based on normal operation records and abnormal simulation records during equipment installation and commissioning, so that the normal operation records cannot simultaneously meet the corresponding numerical conditions and duration conditions, while the continuous abnormal records can simultaneously meet the corresponding conditions.

[0079] When generating any preliminary anomaly record, the corresponding electrical anomaly, temperature anomaly, smoke anomaly, gas anomaly, or image anomaly is written into the anomaly type; the charging location identifier carried by the sensing data involved in the anomaly judgment is written into the preliminary anomaly record; the sampling time when the corresponding anomaly first continuously meets the generation conditions is written into the occurrence time, and the time for which the corresponding anomaly continuously meets the generation conditions is written into the duration. Thus, each preliminary anomaly record includes anomaly type, occurrence time, duration, and charging location identifier, which helps to clearly identify the type, location, and duration of the anomaly.

[0080] By checking the numerical status, trend of change and duration separately, only stable abnormal changes generate preliminary abnormal records. This helps to further eliminate sudden changes in a single sampling point, short-term environmental disturbances and short-term image changes, so that each preliminary abnormal record can clearly represent an actual abnormal phenomenon, thereby reducing erroneous linkage caused by instantaneous fluctuations of a single sensor.

[0081] Then, temporal correlation is performed on the preliminary anomaly records to obtain anomaly evolution chain data. Temporal correlation refers to determining whether multiple anomalies belong to the same continuous development process based on the occurrence order and time interval of the preliminary anomaly records. This is beneficial for adapting to actual situations where current, temperature, smoke, gas, and image response speeds differ. Specific steps include:

[0082] The preliminary abnormal records are arranged to obtain an abnormal record sequence. Specifically, the abnormal judgment unit arranges the preliminary abnormal records of the same charging position according to the time of occurrence to obtain an abnormal record sequence, which can clarify the occurrence order of different types of abnormalities and help avoid mixing abnormal records from different charging positions and those with large time intervals into the same abnormal process.

[0083] The abnormal record sequence is subjected to a correlation determination to obtain abnormal evolution chain data. Specifically, during equipment installation and commissioning, the allowable time interval between different abnormal types is determined based on the sampling period of each sensing device and the response interval of adjacent abnormal phenomena during abnormal simulation. The allowable time interval represents the maximum time after the occurrence of a previous abnormality that can still be considered as part of the same abnormal development process. For example, if an electrical abnormality is detected during abnormal simulation, and a temperature abnormality occurs no later than 25 seconds, and the sampling period of the temperature sensor is 5 seconds, then the allowable time interval between the electrical abnormality and the temperature abnormality is set to 30 seconds. The abnormality judgment unit subtracts the occurrence time of the previous preliminary abnormality record from the occurrence time of the subsequent preliminary abnormality record to obtain the abnormal time interval.

[0084] When a subsequent preliminary anomaly record occurs later than a previous preliminary anomaly record, the anomaly type conforms to the preset anomaly development order, and the anomaly time interval does not exceed the corresponding allowable time interval, a sequential relationship is established. Multiple preliminary anomaly records with sequential relationships form a multi-node anomaly evolution chain; preliminary anomaly records without sequential relationships with other records form a single-node anomaly evolution chain. Anomaly evolution chain data includes anomaly type, occurrence order, and duration; multi-node anomaly evolution chain data also includes the anomaly time interval between adjacent preliminary anomaly records.

[0085] This implementation method can both preserve the case where only a single early anomaly occurs and determine whether multiple anomalies belong to the same continuous development process, which helps to avoid directly identifying a single anomaly as a serious anomaly that has already expanded.

[0086] For example, if a momentary load switch causes a brief increase in charging current, but the current duration does not reach the current duration threshold, no preliminary electrical anomaly record is generated. If the charging current remains abnormal and meets the conditions for generating a preliminary electrical anomaly record, but no other type of preliminary anomaly record is generated within the allowed time interval, a single-node anomaly evolution chain containing only electrical anomalies is generated and identified as an anomaly precursor stage, in which only actions such as reducing charging power and increasing sampling frequency are performed.

[0087] In another abnormal process, after a sustained abnormal charging current, a continuous rise in temperature and a localized high-temperature region appeared sequentially. The initial records of electrical, temperature, and image anomalies conformed to the preset anomaly development sequence, and the time interval between adjacent records did not exceed the corresponding allowable time interval. The system connected the three types of records to form a multi-node anomaly evolution chain. This processing is helpful in distinguishing between instantaneous current fluctuations, single abnormal symptoms, and gradually developing true anomalies.

[0088] Step S02: Perform stage identification on the abnormal evolution chain data to obtain risk stage data, and arrange it with the equipment status data to obtain equipment action sequence data.

[0089] S02.1, Perform stage identification on the abnormal evolution chain data to obtain risk stage data;

[0090] The anomaly judgment unit reads the anomaly type, occurrence order, anomaly time interval, and duration from the anomaly evolution chain data, and determines the risk stage based on the number of records and the combination of anomaly types in the anomaly evolution chain.

[0091] When the anomaly evolution chain is a single-node anomaly evolution chain and includes at least one of the persistent preliminary electrical anomaly records and preliminary temperature anomaly records, risk stage data for recording the anomaly precursor stage is generated.

[0092] The anomaly evolution chain is a multi-node anomaly evolution chain, and includes at least one correlation between the occurrence of an electrical preliminary anomaly record and the occurrence of a temperature preliminary anomaly record and the occurrence of an image preliminary anomaly record, thereby generating risk stage data for the anomaly development stage; the anomaly time interval between adjacent records is used to confirm that the corresponding correlation is still within the allowable time interval.

[0093] When at least one of the persistent preliminary smoke anomaly record and preliminary gas anomaly record appears in the anomaly evolution chain, risk stage data for the anomaly confirmation stage is generated; when the anomaly evolution chain simultaneously meets multiple stage conditions, the risk stage is determined in the order of anomaly confirmation stage, anomaly development stage, and anomaly precursor stage.

[0094] This approach progressively increases the risk stage based on the type of anomaly, its sequence of occurrence, and its duration. This helps to differentiate between different states, such as a single early sign, the continuous expansion of multiple anomalies, and the presence of smoke and gas anomalies, thus avoiding the application of the same treatment intensity to anomalies of different degrees.

[0095] S02.2, Perform action orchestration on the risk stage data and equipment status data to obtain equipment action sequence data; wherein, action orchestration refers to first determining candidate equipment actions based on the risk stage, then using equipment status data to filter executable candidate equipment actions, and determining the execution order according to the safety dependencies between equipment actions, which helps to match equipment actions with the degree of abnormality and the status of on-site equipment.

[0096] Based on the risk stage data, candidate device actions are determined. Specifically, the linkage execution unit reads the risk stage data. In the anomaly precursor stage, the output power of the target charging pile is reduced and the data acquisition frequency of the target charging location is increased. In the anomaly development stage, the data acquisition frequency is increased, the output of the target charging pile is stopped, the target charging circuit is disconnected, and the audible and visual alarm is activated. In the anomaly confirmation stage, the target charging pile is kept stopped from outputting and the target charging circuit is disconnected, the audible and visual alarm is maintained, and the evacuation indicator equipment and ventilation equipment are activated. Thus, the candidate device actions corresponding to the current risk stage are obtained.

[0097] When the risk level rises, retain the equipment actions that still need to be performed in the previous risk level and add the equipment actions corresponding to the current risk level. If the reduction of the target charging pile output power has not been completed and the risk level has risen to the abnormal development stage, cancel the power reduction action and directly stop the output of the target charging pile. This helps to avoid repeatedly sending control information with similar effects to the same equipment.

[0098] The device status data is used to filter and arrange candidate device actions to obtain device action sequence data. Specifically, the linkage execution unit reads the device status data corresponding to each candidate device action. When the target device is online and no fault is recorded, the corresponding candidate device action is retained; when either the target device is offline or a device fault occurs, the corresponding candidate device action is recorded as unexecuted and is not written into the device action sequence data.

[0099] Subsequently, the candidate device actions are arranged according to the safety dependencies between the device actions. Stopping the output of the target charging pile is placed before disconnecting the target charging circuit; the audible and visual alarm action is executed after the abnormal development stage is established, without requiring the target charging circuit to be disconnected as a condition for execution; the evacuation instruction action and the ventilation action are executed after the abnormal confirmation stage is established, without requiring the target charging circuit to be disconnected as a condition for execution. This ensures that actions with safety dependencies are executed sequentially, and that independent safety actions such as alarms, ventilation, and evacuation are executed in a timely manner, which helps to avoid disordered device linkage and the interruption of other safety actions by local device failures.

[0100] The linkage execution unit reads the successful execution records generated during installation, commissioning, and normal operation of the target device when it performs the corresponding action. It then subtracts the control information transmission time from the action completion feedback time to obtain the normal response duration. The maximum value among the various normal response durations is selected, and a communication tolerance time is added to obtain the execution time limit for the corresponding device action.

[0101] The communication tolerance time is determined by the difference between the maximum round-trip time and the average round-trip time of communication test messages during installation and commissioning. For example, if the maximum normal response time for the target device to perform a corresponding action is 3 seconds, and the communication tolerance time is 1 second, then the execution time limit is 4 seconds. This setting helps to cover the device's action time and normal communication fluctuations, avoiding misjudging normal execution as a timeout.

[0102] Subsequently, the target device identifier, device actions, execution order, execution time limit, and completion conditions are written into the same record to obtain device action sequence data.

[0103] The completion condition for reducing output power is that the actual output power returned by the charging pile is not higher than the target power; the completion condition for increasing the data acquisition frequency is that the current sampling period returned by the sensing device is not greater than the target sampling period; the completion condition for stopping the charging pile output is that the charging pile returns to the stop output state; the completion condition for disconnecting the charging circuit is that the charging circuit switching device returns to the disconnect state; the completion conditions for audible and visual alarms, ventilation, and evacuation indications are that the corresponding devices return to the start state. These completion conditions ensure that each device action has a verifiable actual status, which helps avoid judging the completion of a device action solely based on the successful transmission of control information.

[0104] For example, when preliminary electrical anomaly records, preliminary temperature anomaly records, and preliminary image anomaly records form an anomaly evolution chain, the risk stage data records the anomaly development stage. Equipment action sequence data records stopping the target charging pile's output, disconnecting the target charging circuit, and activating the audible and visual alarm. Notably, disconnecting the target charging circuit is executed after the target charging pile stops outputting; the audible and visual alarm action is not contingent on the charging circuit being disconnected. Subsequently, when a preliminary smoke anomaly record is generated within the allowed time interval, the risk stage data is updated to the anomaly confirmation stage, and the equipment action sequence data adds the activation of evacuation indicator equipment and ventilation equipment. This process gradually increases the linkage intensity as the on-site anomaly develops.

[0105] Step S03: Send control information based on the device action sequence data to cause the target device to perform the actions in the device action sequence data.

[0106] Specifically, the linkage execution unit reads the current device actions according to the execution order in the device action sequence data, writes the target device identifier, device action, and action number into the control information, and then sends it to the target device via the Internet of Things communication network. The target device uses existing communication protocols to identify the control information and execute the corresponding actions.

[0107] This implementation method can realize equipment actions without changing the internal structure of the field equipment, which is beneficial for transforming data processing results into actual equipment control.

[0108] Step S04: Obtain feedback data after the device is executed, and adjust the unexecuted device actions according to the feedback data to obtain device execution status data; wherein, the feedback data is used to reflect the actual action status of the target device after receiving control information, and adjusting the unexecuted device actions refers to determining at least one of the following states of the unexecuted device actions: continue execution, wait, block execution, and cancel execution, based on the actual completion status of the previous device actions. This helps to avoid directly identifying the successful transmission of control information as the completion of the on-site action.

[0109] S04.1, Perform an action completion determination on the feedback data to obtain the action completion result;

[0110] Specifically, after receiving the control information, the target device returns feedback data, which includes the target device identifier, action number, feedback time, actual device status, and fault status. The linkage execution unit reads the sending time and execution time limit of the corresponding control information and verifies the target device identifier and action number.

[0111] If control information has been sent but corresponding feedback data has not yet been received, and the time between the current moment and the time the control information was sent has not exceeded the execution time limit, it is recorded as sent status; if feedback data has been received, but the actual equipment status has not yet met the completion conditions, and no fault status is recorded, and the execution time limit has not been exceeded, it is recorded as executing status; if the actual equipment status meets the completion conditions, it is recorded as action completed status; if the execution time limit has been exceeded and the completion conditions are still not met, or if any of the following situations occurs: feedback data records equipment fault, it is recorded as execution abnormal status, and the action completion result is obtained.

[0112] This process distinguishes the device action status based on feedback reception, actual device status, and execution time limit, which helps to avoid mistakenly believing that the device action has been completed when control information has been sent.

[0113] S04.2, adjust the unexecuted device actions based on the action completion results to obtain device execution status data;

[0114] Specifically, when an action is completed, the linkage execution unit writes the current device action into the completed record and enables the next device action that depends on the completion of the current device action. If the action completion result record is in either the sent or in-process state, the corresponding dependent action remains in a waiting state. This process helps prevent the next device action from starting prematurely before the previous device action is completed.

[0115] Specifically, firstly, execution anomalies in the action completion results are identified to obtain abnormal action data; when an execution anomaly is recorded in the action completion result, the linkage execution unit extracts the target device identifier, action number, and anomaly cause to obtain abnormal action data. The anomaly cause includes at least one of execution timeout and equipment failure, which helps to accurately map the execution anomaly to a specific device and specific action.

[0116] Then, based on the abnormal action data, the actions of the unexecuted equipment are adjusted to obtain equipment execution status data. After the abnormal action data is formed, the linkage execution unit searches for unexecuted equipment actions in the equipment action sequence data according to the action number corresponding to the abnormal action, and determines whether each unexecuted equipment action depends on the abnormal action for completion. Then, it combines the risk stage data and equipment status data to determine the execution status of the corresponding unexecuted equipment action. The specific steps include:

[0117] The dependent actions corresponding to the abnormal action data are blocked to obtain the action data to be adjusted. Specifically, the linkage execution unit writes all unexecuted device actions into the action data to be adjusted, marks unexecuted device actions that depend on the abnormal actions as blocked, and records the corresponding device actions, target device identifiers, and dependencies. This process only blocks subsequent actions that do depend on the abnormal actions, which helps to avoid local execution anomalies affecting other independent safety actions.

[0118] By combining the risk stage data and equipment status data, the action data to be adjusted is obtained to obtain the equipment execution status data. Specifically, the linkage execution unit reads the risk stage data, determines whether the action to be adjusted still matches the current risk stage, and then reads the online status and fault status of the corresponding target equipment.

[0119] When the action to be adjusted does not match the current risk stage, it is recorded as canceled execution status; when the action to be adjusted depends on the completion of an abnormal action, it remains blocked; when the action to be adjusted does not depend on the completion of an abnormal action and the target device is online and no fault is recorded, it is recorded as continue execution status; when either the target device is offline or the device is faulty, it is recorded as execution abnormal status.

[0120] Finally, the linkage execution unit collects the sent status, execution status, action completion status, execution abnormal status, continue execution status, blockage status, and canceled execution status to obtain the equipment execution status data. This helps to limit the impact of equipment execution abnormalities and allows independent safety actions such as alarms, ventilation, and evacuation to continue to be executed.

[0121] For example, after the target charging pile returns to the stop output state, the linkage execution unit sends control information to disconnect the charging circuit. If the charging circuit switching device fails to return to the disconnected state within the execution time limit, the disconnection action is recorded as an execution anomaly, and the actions of unexecuted equipment that actually depend on the charging circuit disconnection state are blocked. Audible and visual alarms, ventilation, and evacuation indications do not depend on the charging circuit disconnection and continue to execute when the corresponding risk stage is reached, while outputting charging circuit disconnection anomaly information. This helps to prevent the processing terminal from mistakenly believing that the power has been cut off on site and prevents a single equipment failure from blocking other safety procedures.

[0122] Step S05: Obtain the sensing data after the device operates, perform recovery status analysis on the sensing data, and obtain recovery status data.

[0123] Specifically, after the equipment performs its action, the data processing unit continues to acquire new charging electrical data from the charging pile controller, new environmental perception data from the environmental sensors, and new image perception data from the image acquisition device to obtain the perception data after the equipment action. The perception data carries the acquisition time and the target charging location identifier, which helps to accurately reflect the state changes of the target charging location after the equipment action.

[0124] Recovery status analysis refers to re-performing timing validity analysis, abnormal change extraction, and timing correlation on the sensing data after equipment operation, and then combining it with equipment execution status data to determine whether the site has continued to recover.

[0125] For temperature, smoke concentration, gas concentration, and image sensing data, the judgment boundaries, duration thresholds, and temporal correlation rules from the anomaly identification phase are retained. For charging electrical data directly altered by reduced charging power, stopped charging output, or disconnected charging circuits, the baseline current before the device action is executed is no longer used to determine the recovery status. Instead, the completion conditions and feedback data in the device action sequence data are used to determine whether the charging pile and charging circuit have reached the corresponding safe state. This approach helps avoid misjudging proactive current reduction and proactive power disconnection as persistent electrical anomalies.

[0126] During equipment installation and commissioning, for each risk stage, the time required for the temperature, smoke concentration, gas concentration, and image abnormality to stop and remain stable after the equipment action is completed is recorded, thus obtaining the environmental stabilization time. The recovery control unit determines the maximum sampling period from the sampling periods of each sensing device and takes the larger value between the maximum sampling period and the corresponding environmental stabilization time as the recovery duration.

[0127] Once the charging pile and charging circuit reach the safe state defined by the equipment action sequence data, and all preliminary abnormality records (temperature, smoke, gas, and image) disappear, determine the moment when the aforementioned preliminary abnormality records last disappeared, and start accumulating the recovery observation time from that moment. Subtract the moment when the last disappeared from the current moment to obtain the current recovery observation time.

[0128] If the current recovery observation time reaches the recovery duration, no abnormal evolution chain data is re-formed during the recovery observation period, and no abnormal execution of key safety actions is recorded in the equipment execution status data, recovery status data indicating stable recovery on site is generated.

[0129] If any of the following conditions persists: initial temperature anomaly record, initial smoke anomaly record, initial gas anomaly record, or initial image anomaly record; during the recovery observation period, the anomaly evolution chain data is re-formed; or if any of the following conditions occur: key safety actions have not yet been completed, recovery status data indicating continued linkage is generated.

[0130] This implementation first confirms whether the charging device has reached a controlled safety state, then confirms whether the environmental anomaly has disappeared, then confirms whether the anomaly disappearance state continues, and finally confirms whether the anomaly development process has reappeared. This helps to avoid incorrect judgments caused by active power-off and reduces the repeated starting and stopping of the device near the judgment boundary.

[0131] Step S06: Perform linkage status control based on the recovery status data and device execution status data to obtain linkage control results.

[0132] Specifically, linkage status control refers to determining at least one control state among maintaining, releasing, and restoring normal settings for already started equipment actions based on whether the site has stabilized and whether the actions of each device have actually been completed.

[0133] The recovery control unit reads recovery status data and equipment execution status data. While the recovery status data recording continues to maintain linkage, it sends hold control information to the already activated target equipment, ensuring that the reduced charging power, increased data acquisition frequency, audible and visual alarms, ventilation, evacuation indication, and charging circuit disconnection status are maintained. This process helps maintain on-site handling even when the anomaly has not yet stabilized and critical safety actions have not been completed.

[0134] Once the recovery status data records that the site has been stably restored and the equipment execution status data indicates that the critical safety actions required for the current risk stage have been completed, the linkage status is lifted based on the highest risk stage reached by this anomaly.

[0135] If the anomaly only reaches the pre-anomaly stage and the actions of stopping charging output and disconnecting the charging circuit are not executed, the recovery control unit will restore the sampling period of the sensing device to the normal sampling period and restore the output power of the target charging pile to the set value before the anomaly occurred.

[0136] If the anomaly reaches either the anomaly development stage or the anomaly confirmation stage, the recovery control unit will send a release control message to the audible and visual alarms and evacuation indicator devices, and then send a stop control message to the ventilation equipment. The sampling cycle of the sensing devices will also be restored to the normal sampling cycle. Simultaneously, the target charging pile will remain stopped and the target charging circuit will be disconnected. Whether the target charging pile and target charging circuit should be restored will be determined separately by on-site personnel after a safety inspection; they will not be automatically reconnected by the current linkage status control.

[0137] After the target device performs actions to maintain, release, and restore normal settings, it returns the final feedback data. The recovery control unit collects the output status of the target charging pile, the sampling period of each sensing device, the on / off status of the charging circuit switching device, the alarm status of the audible and visual alarm, the operating status of the ventilation equipment, and the indication status of the evacuation indicator device according to the target device identifier, and obtains the linkage control result.

[0138] The linkage control results are used to indicate the actual status of each device at the current charging location: whether the linkage is maintained, whether the linkage is deactivated, and whether the safety isolation is maintained. This is beneficial for continuing to handle the situation when the anomaly has not yet been stabilized and disappeared, and for orderly removing unnecessary device actions after the on-site status has been continuously restored.

[0139] For example, after the initial abnormal smoke record appears in the abnormal evolution chain data, during the risk stage data recording abnormality confirmation stage, the target charging pile, charging circuit switching device, audible and visual alarm, ventilation equipment, and evacuation indicator equipment execute corresponding actions according to the equipment action sequence data. After linkage, the charging pile stops outputting and the charging circuit remains disconnected. Temperature, smoke concentration, gas concentration, and image anomalies continuously disappear, and if the abnormal evolution chain data is not re-formed within the recovery period, the recovery control unit deactivates the audible and visual alarm and evacuation indicator, stops the ventilation equipment, restores the normal sampling cycle of the sensing equipment, and continues to keep the target charging pile stopped outputting and the charging circuit disconnected. This process helps avoid repeated start-stop of alarm, ventilation, and evacuation equipment and maintains critical safety isolation before the cause of the anomaly has been investigated on-site.

[0140] This embodiment first distinguishes between instantaneous disturbances and continuously developing real anomalies through anomaly timing analysis. Then, through stage identification and action orchestration, it gradually increases the equipment actions as the anomaly develops. During equipment execution, the actual feedback from the target equipment determines whether the linkage process should continue, and it blocks and adjusts unexecuted equipment actions when equipment execution malfunctions. After linkage is completed, continuous recovery judgment determines whether to maintain or deactivate the equipment linkage. This complete process can reduce erroneous linkages caused by short-term disturbances, shorten the confirmation time of real anomalies, and ensure that equipment actions are consistent with the actual on-site state.

[0141] like Figure 2 As shown, the present invention is superior to the prior art in terms of the timeliness of anomaly identification, the ability to suppress false triggers, the accuracy of linkage execution, and the stability of recovery control. The present invention can identify real anomalies that develop continuously, reduce false triggers caused by instantaneous fluctuations, and adjust subsequent actions according to the actual feedback from the equipment; at the same time, combined with the linkage release of continuous recovery status control, it is beneficial to improve the accuracy and stability of linkage handling.

[0142] As a second aspect of this embodiment, this embodiment also provides an IoT device linkage control system based on multimodal timing triggering, including:

[0143] The data processing unit is used to acquire sensing data and device status data, perform anomaly time series analysis on the sensing data, and obtain anomaly evolution chain data.

[0144] An anomaly detection unit is used to identify the stage of the anomaly evolution chain data, obtain risk stage data, and arrange it with equipment status data to obtain equipment action sequence data.

[0145] The linkage execution unit is used to send control information based on the device action sequence data, so as to cause the target device to perform the actions in the device action sequence data; to obtain feedback data after the device execution, and to adjust the unexecuted device actions according to the feedback data to obtain device execution status data;

[0146] The recovery control unit is used to acquire sensing data after the device operates, perform recovery status analysis on the sensing data to obtain recovery status data, and perform linkage status control based on the recovery status data and the device execution status data to obtain linkage control results.

[0147] The embodiments of the present invention described above are subject to modification and change of method by those skilled in the art without departing from the embodiments and broader aspects of the present invention. The appended claims are intended to include all such modifications and changes of method that do not depart from the present invention.

[0148] Those skilled in the art should understand that the embodiments of the present invention can be implemented using a pure hardware architecture, a pure software architecture, or an integrated hardware and software architecture. The present invention can be prepared as a computer program product, which can be stored in various non-volatile computer-readable storage media, including but not limited to solid-state drives, flash memory chips, mobile storage devices, optical discs, cloud storage servers, and other standardized storage media, and is not limited to traditional storage media.

[0149] Based on the foregoing description in conjunction with the accompanying drawings, those skilled in the art will understand that the embodiments of this application can also be implemented by software programs. Therefore, this application also provides a computer-readable storage medium. This computer-readable storage medium stores computer-readable instructions thereon, which, when executed by one or more processors, implement the method described above in conjunction with the accompanying drawings.

[0150] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0151] It should be noted that although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the steps depicted in the flowchart can be performed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0152] It should be understood that when the terms "first," "second," "third," and "fourth," etc., are used in the claims, specification, and drawings of this application, they are used only to distinguish different objects and not to describe a specific order. The terms "comprising" and "including" as used in the specification and claims of this application indicate the presence of the described features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof.

[0153] It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application. As used in this specification and claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this specification and claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations.

[0154] Although the embodiments of this application are described above, the content is merely an example adopted for the purpose of facilitating understanding of this application and is not intended to limit the scope and application scenarios of this application. Any person skilled in the art described in this application may make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in this application, but the scope of patent protection of this application shall still be determined by the scope defined in the appended claims.

Claims

1. A linkage control method for Internet of Things devices based on multimodal time series triggering, characterized in that, include: Acquire sensing data and device status data, perform anomaly timing analysis on the sensing data, and obtain anomaly evolution chain data; The abnormal evolution chain data is subjected to stage identification to obtain risk stage data, which is then combined with the equipment status data to obtain equipment action sequence data. Control information is sent based on the device action sequence data to cause the target device to perform the actions in the device action sequence data. Obtain feedback data after the device executes the action, and adjust the actions of the unexecuted device based on the feedback data to obtain the device execution status data; Acquire sensing data after device action, perform recovery status analysis on the sensing data, and obtain recovery status data; Based on the recovery status data and the device execution status data, linkage status control is performed to obtain the linkage control result.

2. The IoT device linkage control method according to claim 1, characterized in that, Anomaly time-series analysis is performed on the perceived data to obtain anomaly evolution chain data, including: Perform time series validity analysis on the sensed data to obtain reliable time series data; Anomaly identification is performed on the trusted time-series data to obtain anomaly evolution chain data.

3. The IoT device linkage control method according to claim 2, characterized in that, Anomaly identification is performed on the trusted time-series data to obtain anomaly evolution chain data, including: Anomaly extraction is performed on the trusted time-series data to obtain preliminary anomaly records; Perform time-series correlation on the preliminary anomaly records to obtain anomaly evolution chain data.

4. The IoT device linkage control method according to claim 3, characterized in that, Perform time-series correlation on the preliminary anomaly records to obtain anomaly evolution chain data, including: The preliminary abnormal records are arranged to obtain an abnormal record sequence; Perform a correlation determination on the abnormal record sequence to obtain abnormal evolution chain data.

5. The IoT device linkage control method according to claim 1, characterized in that, The abnormal evolution chain data is subjected to stage identification to obtain risk stage data, which is then combined with equipment status data to obtain equipment action sequence data, including: Stage identification is performed on the abnormal evolution chain data to obtain risk stage data; The risk stage data and equipment status data are processed to perform action orchestration, resulting in equipment action sequence data.

6. The IoT device linkage control method according to claim 5, characterized in that, The risk phase data and equipment status data are used to perform action orchestration to obtain equipment action sequence data, including: The candidate device actions are determined based on the aforementioned risk stage data; The device status data is used to filter candidate device actions and arrange them to obtain device action sequence data.

7. The IoT device linkage control method according to claim 1, characterized in that, Adjusting unexecuted device actions based on feedback data yields device execution status data, including: The feedback data is used to determine the completion of the action and obtain the action completion result. Adjust the unexecuted device actions based on the completion results of the actions to obtain device execution status data.

8. The IoT device linkage control method according to claim 7, characterized in that, Based on the completion result of the action, adjust the actions of the unexecuted devices to obtain device execution status data, including: Identify execution anomalies in the completion results of the actions to obtain abnormal action data; Based on the abnormal action data, the actions of the unexecuted devices are adjusted to obtain the device execution status data.

9. The IoT device linkage control method according to claim 8, characterized in that, Based on the abnormal action data, the actions of the unexecuted devices are adjusted to obtain device execution status data, including: Block the dependent actions corresponding to the abnormal action data to obtain the action data to be adjusted; By combining the risk stage data and equipment status data, the action data to be adjusted is obtained to obtain the equipment execution status data.

10. A multimodal timing-triggered IoT device linkage control system, applicable to the IoT device linkage control method according to any one of claims 1 to 9, characterized in that, include: The data processing unit is used to acquire sensing data and device status data, perform anomaly time series analysis on the sensing data, and obtain anomaly evolution chain data. An anomaly detection unit is used to identify the stage of the anomaly evolution chain data, obtain risk stage data, and arrange it with equipment status data to obtain equipment action sequence data. The linkage execution unit is used to send control information based on the device action sequence data, so as to cause the target device to perform the actions in the device action sequence data; to obtain feedback data after the device execution, and to adjust the unexecuted device actions according to the feedback data to obtain device execution status data; The recovery control unit is used to acquire sensing data after the device operates, perform recovery status analysis on the sensing data to obtain recovery status data, and perform linkage status control based on the recovery status data and the device execution status data to obtain linkage control results.

Images