An explosion-proof multi-sensor intelligent linkage control method

Abstract

The present application relates to the technical field of industrial safety monitoring and control, and discloses an explosion-proof multi-sensor intelligent linkage control method, comprising: an explosion-proof wireless data transceiver collects temperature, combustible gas and smoke data and calculates a rapid prediction value; when the limit is exceeded, a temporary emergency linkage instruction with an event identifier is sent to a programmable controller through an independent interrupt channel; the programmable controller extracts complete cycle data of the ordinary data channel to calculate a refined fusion value, and executes instruction review in combination with confirmation and cancellation thresholds; the controller switches between simplified weighted fusion and complete Kalman filtering mode based on the change rate of monitoring parameters, generates an adaptive alarm threshold based on real-time noise level for multi-level linkage control, and performs spatial cross verification in combination with adjacent node support markers. The present application overcomes technical difficulties such as false alarms of fixed thresholds and slow response of network congestion, and takes into account both extremely early rapid warning and high reliability of multi-level linkage.

Description

Technical Field

[0001] This invention relates to the field of industrial safety monitoring and control technology, and in particular to an explosion-proof multi-sensor intelligent linkage control method. Background Technology

[0002] In explosion-proof environments such as petrochemical plants, mines, and oil depots, rapid and reliable detection of early fires or flammable gas leaks is crucial for ensuring personal and property safety. Existing monitoring solutions typically employ fixed threshold alarms: each sensor independently determines whether a threshold is exceeded, and then a central controller processes the data. However, fixed thresholds struggle to balance rapid response with false alarm control: thresholds that are too low can lead to frequent triggering by noise and transient disturbances, resulting in false alarms; thresholds that are too high can delay response to actual hazards.

[0003] Furthermore, traditional solutions often rely on a central controller for unified fusion and decision-making. Sensor data needs to be uploaded periodically through ordinary communication channels, which introduces delays when the data volume is large or the network is congested, making it difficult to meet the response requirements for the early and rapid development of explosive hazards. At the same time, existing technologies lack effective mechanisms to suppress single-point false triggering and also lack means to dynamically adjust the fusion strategy according to changes in operating conditions, making it difficult to achieve a stable balance between abrupt and stable operating conditions.

[0004] Therefore, an explosion-proof multi-sensor intelligent linkage control method is needed, which enables edge nodes to quickly trigger central pre-linkage through an independent interrupt channel while ensuring complete data upload. The central node then performs fine-grained confirmation or cancellation after receiving complete data, and reduces false triggering and improves response reliability by using spatial cross-validation and dynamic fusion strategies. Summary of the Invention

[0005] This invention aims to at least solve one of the technical problems existing in related technologies. To this end, this invention provides an explosion-proof multi-sensor intelligent linkage control method.

[0006] An explosion-proof multi-sensor intelligent linkage control method is applied to a system comprising multiple explosion-proof wireless data transceivers and a programmable controller, wherein each of the explosion-proof wireless data transceivers is configured with a corresponding set of adjacent nodes, including: The explosion-proof wireless data transceiver collects temperature data, combustible gas concentration data, and smoke concentration data, and calculates a rapid prediction value based on the temperature data, combustible gas concentration data, and smoke concentration data. When any of the temperature data, the combustible gas concentration data, or the smoke concentration data exceeds the corresponding rapid threshold, or when the rapid prediction value is not lower than the prediction threshold, the explosion-proof wireless data transceiver first sends a temporary emergency linkage command with an event identifier to the programmable controller through an independent interrupt channel, and periodically uploads the temperature data, the combustible gas concentration data, and the smoke concentration data through a normal data channel. After receiving the temporary emergency linkage command, the programmable controller extracts the temperature data, combustible gas concentration data and smoke concentration data received from the normal data channel in the most recent complete sampling period according to the event identifier, and calculates the refined fusion value; The programmable controller compares the refined fusion value with preset confirmation and cancellation thresholds to perform actions such as maintaining, upgrading, or canceling the temporary emergency linkage command.

[0007] Furthermore, the calculation of the rapid prediction value based on the temperature data, the combustible gas concentration data, and the smoke concentration data includes: The temperature data, the combustible gas concentration data, and the smoke concentration data are respectively subjected to minimum-maximum normalization. The membership degree of the normalized temperature data, combustible gas concentration data, and smoke concentration data is calculated and output using a piecewise linear membership function. The maximum value of the three membership degree outputs is taken as the fast prediction value; the calculation formula for the fast prediction value is as follows: in: Let be the rapid prediction value of the i-th explosion-proof wireless data transceiver at time t, which is dimensionless; For temperature data, This is data on combustible gas concentration. Data on smoke concentration; The temperature membership function, Let f be the membership function of combustible gas concentration. Let f be the membership function of smoke concentration. ( ), ( ), ( ) are the piecewise linear membership functions output after normalization of the corresponding sensor data.

[0008] Furthermore, the calculation of the refined fusion value includes calculating the refined fusion value using a weighted summation method; The programmable controller compares the refined fusion value with preset confirmation and cancellation thresholds respectively to perform actions such as maintaining, upgrading, or canceling the temporary emergency linkage command, including: When the refined fusion value is not lower than the confirmation threshold, a danger is confirmed, and the temporary emergency linkage instruction is maintained or upgraded. When the refined fusion value is lower than the revocation threshold, no danger is confirmed, and a revocation command is sent to the corresponding explosion-proof wireless data transceiver and linkage execution terminal. When the refined fusion value is between the cancellation threshold and the confirmation threshold, the system maintains the linkage state of the previous moment and triggers the continuous refined fusion task of the next cycle.

[0009] Furthermore, the method also includes a step of detecting sudden changes in operating conditions and switching modes, specifically: The programmable controller calculates the rate of change of the monitoring parameters based on the temperature data, the combustible gas concentration data, and the smoke concentration data in adjacent sampling periods; The rate of change of the monitored parameters is compared with the dynamic rate of change threshold. When the rate of change of the monitored parameter exceeds the dynamic rate of change threshold, the simplified data fusion mode is switched and the fusion output value is calculated as the current monitored parameter. At the same time, a high sampling frequency status command is sent to the corresponding explosion-proof wireless data transceiver. When the simplified data fusion mode is in effect, and the rate of change of the monitored parameter is continuously lower than the preset recovery threshold and maintains a preset recovery time, the sensor fusion estimate is calculated using the full Kalman filter mode as the current monitored parameter; the recovery threshold is lower than the dynamic rate of change threshold.

[0010] Furthermore, in: Let be the rate of change of the monitoring parameters of the i-th node at time t. The temperature data is at time t. This is the temperature data from the previous sampling time. This is the preset range reference value for temperature data. Here is the data on the concentration of combustible gas at time t. This is the combustible gas concentration data from the previous sampling time. This is the preset range reference value for combustible gas concentration. Here is the smoke concentration data at time t. This is the smoke concentration data from the previous sampling time. This is the preset range reference value for smoke concentration. The sampling interval is... The formula for calculating the dynamic change rate threshold is as follows: in: The threshold for the rate of change at time t; The average rate of change within the historical rate of change window; The standard deviation of the rate of change within the historical rate of change window; This is the proportionality coefficient.

[0011] Furthermore, the switching to simplified data fusion mode and calculating the fusion output value as the current monitoring parameter includes: Only the data from the two most recent sampling periods are retained, and an adaptive weighted average of the rate of change is performed on each sensor to obtain the fused output value. The calculation formula is as follows: in: Let be the fused output value of the i-th node at time t. This is the current sampled value. The previous sampled value, Let be the adaptive weights at time t. Within the range [0,1]; As a simplified data fusion model output, it is used for over-limit judgment and linkage confirmation; The formula for calculating the adaptive weights is as follows: in: Let be the adaptive weights at time t. To preset the minimum weight, To preset the maximum weight, The rate of change; The threshold for the rate of change. Greater than zero; and Within the interval (0,1) and Less than .

[0012] Furthermore, the method also includes an adaptive threshold generation step based on the real-time noise level, specifically: The programmable controller calculates the short-term standard deviation of the historical data of each sensor within the first time window. Based on the corresponding base threshold and the short-term standard deviation, adaptive alarm thresholds are generated for each sensor; the calculation formula for the adaptive alarm threshold is as follows: in: Let be the adaptive alarm threshold for sensor j at time t. The basic threshold of sensor j, This is the proportionality coefficient. Let j be the short-term standard deviation of sensor j within the first time window. Greater than zero; Non-negative; Greater than zero; It replaces fixed thresholds for exceeding limits and serves as a support marker for spatial cross-validation; The programmable controller uses the short-term standard deviations corresponding to each sensor as the observation noise covariance and inputs them into the complete Kalman filter mode for state updates.

[0013] Furthermore, the method also includes a multi-level linkage execution control step, specifically: The programmable controller compares the current monitoring parameters of each sensor with the first linkage threshold and the second linkage threshold, wherein the first linkage threshold is equal to the corresponding adaptive alarm threshold, and the second linkage threshold is the adaptive alarm threshold multiplied by a preset escalation coefficient. When the current monitoring parameter of any sensor exceeds the first linkage threshold, ventilation and early warning operations are initiated. When the current monitoring parameter continues to exceed the limit, and the current monitoring parameter corresponding to the smoke concentration data and the current monitoring parameter corresponding to the temperature data both exceed their respective second linkage thresholds and remain so for a preset time, the fire extinguishing and audible and visual alarm operations are triggered.

[0014] Furthermore, the method also includes a spatial cross-validation step for temporary emergency response commands, specifically: The programmable controller obtains the current monitoring parameters of each neighboring node in the neighboring node set at the corresponding time. When the current monitoring parameter of a certain neighboring node reaches the corresponding adaptive alarm threshold, the support flag of the neighboring node is set to 1, otherwise it is set to 0. The spatial cross-validation result is calculated by combining the spatial weights between the explosion-proof wireless data transceiver and each adjacent node, as well as the corresponding support tags. The formula for calculating the spatial cross-validation result is as follows: in: Let be the spatial cross-validation result of the event corresponding to the i-th node at time t. Indexing adjacent nodes Let i be the set of neighboring nodes of node i; Let node i and node Spatial weights between them Adjacent nodes Support tag at time t, when node The value is set to 1 when the current monitoring parameter reaches the corresponding adaptive alarm threshold, and 0 otherwise. When the spatial cross-validation result is lower than the spatial confirmation threshold and the refined fusion value is not confirmed as dangerous, the programmable controller sends the revocation command to the corresponding explosion-proof wireless data transceiver and linkage execution terminal.

[0015] Furthermore, the method also includes a control chain time base binding step, specifically: Establish a set of commissioning calibration parameters with version numbers according to node model, sensor type, measurement unit, and monitoring area operating conditions; When generating the temporary emergency linkage command and binding the event identifier, the current version of the commissioning calibration parameter group is locked; the version of the commissioning calibration parameter group is not switched before the event is retained, revoked, upgraded or terminated.

[0016] The above-described one or more technical solutions in the embodiments of the present invention have at least one of the following technical effects: 1. A dual-channel mechanism with both normal and independent interrupts was designed. Edge nodes utilize lightweight fuzzy logic for millisecond-level prediction. Once the limit is exceeded, the high-priority independent interrupt channel is immediately used to preempt transmission, completely solving the fatal delay problem of the inability to send dangerous initial signals in a timely manner under conditions of massive data upload or network congestion.

[0017] 2. It abandons the crude logic of traditional trigger-based action. Through "refined fusion" based on complete cycle data at the central end, a "cancel / upgrade" state machine with a hysteresis comparison interval, and "cross-verification of adjacent nodes" in the physical space dimension, a triple mechanism effectively filters out false alarms caused by local noise, instantaneous dust interference, or single sensor hardware failure.

[0018] 3. A seamless switching technology for filtering modes based on dynamic rate of change was developed. When the environment is stable, Gaussian background noise is removed by relying on full Kalman filtering; when a sudden fire or leak causes a sudden change in data, the long historical window is instantly frozen and switched to simplified weighted fusion that retains only two periods, overcoming the problem of severe data lag in traditional filtering algorithms during dangerous emergencies.

[0019] 4. By extracting the short-term standard deviation of field data in real time, an "adaptive alarm threshold" is generated, ensuring that the system is not overly sensitive under noisy conditions and not sluggish under quiet conditions. This dynamic threshold drives multi-level physical linkages (such as single-point over-limit ventilation and multi-point multi-parameter dual over-limit fire suppression), achieving a precise and safe match between risk levels and physical actions.

[0020] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is the main flowchart of the explosion-proof multi-sensor intelligent linkage control method of the present invention; Figure 2 This is a logic diagram of the judgment and execution of the refined fusion task of the present invention; Figure 3 This is a logic diagram of the fusion mode switching and adaptive weight calculation of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention. The following embodiments are used to illustrate this invention but cannot be used to limit the scope of this invention.

[0024] like Figure 1 The figure shows the main flow of an explosion-proof multi-sensor intelligent linkage control method.

[0025] Example 1: Basic Complete Example This embodiment applies to a system comprising multiple explosion-proof wireless transceivers and a programmable controller. Each explosion-proof wireless transceiver integrates a temperature sensor, a combustible gas sensor, and a smoke sensor, forming a monitoring matrix through a communication network. Each explosion-proof wireless transceiver is configured with a node identifier, a spatial location identifier, and a set of neighboring nodes. The node identifier is a unique integer number, the spatial location identifier is the node's planar coordinates or area number within the monitoring area, and the set of neighboring nodes is established during the commissioning calibration phase based on the spatial location identifier, including the numbers of other nodes whose straight-line distance from the node does not exceed a preset distance threshold.

[0026] In step S1, each explosion-proof wireless transceiver establishes a normal data channel and an independent interrupt channel. The normal data channel periodically uploads temperature data. Combustible gas concentration data and smoke concentration data The sampling period is The independent interrupt channel has a higher transmission priority than the normal data channel and is used to upload fast-predicted events. These fast-predicted events carry event identifiers and node identifiers, and their length is much shorter than a normal data frame. When both the normal data channel and the independent interrupt channel are ready to transmit, data from the independent interrupt channel is transmitted first.

[0027] In step S2, the microcontroller of the explosion-proof wireless data transceiver runs a fast prediction module based on fuzzy inference. Temperature data, combustible gas concentration data, and smoke concentration data are first subjected to min-max normalization, mapping them to the [0,1] interval; then, a piecewise linear membership function is used to calculate the membership output. ( ), ( )and ( The rapid prediction value is calculated according to formula (1). (1) in: Let be the rapid prediction value of the i-th explosion-proof wireless data transceiver at time t, which is dimensionless; For temperature data, This is data on combustible gas concentration. Data on smoke concentration; The temperature membership function, Let f be the membership function of combustible gas concentration. Let f be the membership function of smoke concentration. ( ), ( ), ( These are the piecewise linear membership functions of the corresponding sensor data after normalization, and are dimensionless.

[0028] Temperature data, combustible gas concentration data, and smoke concentration data are first mapped to the [0,1] interval through min-max normalization, and then substituted into the corresponding piecewise linear membership function; the maximum value of the three is taken as the rapid prediction value.

[0029] The system compares the threshold with the rapid threshold and the prediction threshold to determine whether to trigger a temporary emergency response command.

[0030] When any sensor signal in the temperature data, combustible gas concentration data, or smoke concentration data exceeds the corresponding fast threshold or fast prediction value... When the data level is not lower than the predicted threshold, the microcontroller sends a temporary emergency linkage command with an event identifier to the programmable controller via an independent interrupt channel, while continuing to perform complete data acquisition and filtering through the normal data channel. The rapid threshold is 2 to 3 times the upper limit of the normal fluctuation range of the corresponding sensor. The normal fluctuation range is determined by the data acquired by the explosion-proof wireless data transceiver within the safety condition calibration window.

[0031] In step S3, after receiving the temporary emergency linkage command, the programmable controller latches the corresponding temporary linkage state according to the event identifier and initiates the refined fusion task corresponding to the event identifier. The refined fusion task takes temperature data, combustible gas concentration data, and smoke concentration data received from the ordinary data channel within the most recent complete sampling period as input, calculates the membership degree of the three types of data respectively, and weights and fuses them to obtain the refined fusion value. ;when When the level is not lower than the confirmation threshold, confirm the danger and maintain or escalate the temporary emergency response order; when If the threshold for revocation is lower than the hazard is not confirmed, a revocation command is sent to the corresponding explosion-proof wireless transceiver and the linkage execution terminal; when... When the threshold is between the cancellation threshold and the confirmation threshold, the system maintains the linkage state of the previous moment and triggers the continuous fine-tuning fusion task of the next cycle.

[0032] The judgment and execution logic of refined fusion tasks is as follows: Figure 2 As shown.

[0033] Refined fusion value The weighted summation method is used for calculation, and the specific formula is as follows: in, , , These are the fusion weights for temperature, combustible gas, and smoke data, respectively.

[0034] The fusion weights are determined during the commissioning and calibration phase based on the inverse normalization of the measurement variance of each sensor under safe operating conditions, and satisfy the following conditions: .

[0035] Both the confirmation threshold and the cancellation threshold are dimensionless preset constants, whose values ​​are determined by safety operating condition data and hazardous gas diffusion models during the commissioning and calibration phase. To prevent frequent false alarms in the critical state, the confirmation threshold is strictly greater than the cancellation threshold, forming a hysteresis comparison interval (for example, the confirmation threshold ranges from 0.75 to 0.90, and the cancellation threshold ranges from 0.30 to 0.50).

[0036] In step S4, the programmable controller calculates the rate of change of the monitored parameters in real time. The rate of change is calculated according to equation (2). (2) in: Let be the rate of change of the monitoring parameters of the i-th node at time t. The data is the temperature at time t, in degrees Celsius. This refers to the temperature data from the previous sampling time. This is the preset range reference value for temperature data; The data represents the concentration of combustible gas at time t, expressed as a volume percentage. This is the combustible gas concentration data from the previous sampling time; This is the preset range reference value for the concentration of combustible gas; Here is the smoke concentration data at time t, in units of shading rate. This refers to the smoke concentration data at the previous sampling time. This is the preset range reference value for smoke concentration; , , The safety operating condition data is used to determine the conditions during the commissioning and calibration phase. The sampling interval is in seconds.

[0037] The maximum value of the change in temperature, combustible gas concentration, and smoke concentration after normalization to the baseline value within adjacent sampling periods is obtained by dividing by the sampling interval.

[0038] With dynamic rate of change threshold The comparison determines the switching of fusion modes and adaptive weights.

[0039] The dynamic rate of change threshold is determined according to formula (3). (3) in: The threshold for the dynamic rate of change at time t, in units of . same; The average rate of change within the historical rate of change window, in units of and . same; The standard deviation of the rate of change within the historical rate of change window, in units of same; This is a proportionality constant, dimensionless, with a value range of 2 to 3.

[0040] Within a historical rate of change window, the mean and standard deviation of the rate of change are statistically analyzed and overlaid using control chart principles. The threshold is set at 1 / 2 standard deviation.

[0041] The historical window length shall be no less than 30 sampling periods; Non-negative; Greater than zero.

[0042] Used to determine if the current operating conditions have changed abruptly, triggering a simplified data fusion mode.

[0043] When the rate of change Exceeding the dynamic change rate threshold When the change rate is continuously below the recovery threshold and the recovery time is maintained, the programmable controller switches to simplified data fusion mode, retains only the data from the two most recent sampling periods, and performs adaptive weighted averaging of the change rate according to equations (4) and (5); at the same time, it sends a high sampling frequency status command to the corresponding explosion-proof wireless data transceiver to shorten the sampling interval. When the change rate is continuously below the recovery threshold and the recovery time is maintained, the complete Kalman filter and historical data window are restored. The recovery threshold is lower than the dynamic change rate threshold, and the recovery time is not less than 10 sampling periods.

[0044] The logic for fusion mode switching and adaptive weight calculation is as follows: Figure 3 As shown.

[0045] (4) in: This is the fusion output value of the i-th node at time t, with the same dimensions as the original sensor data. This is the current sampled value. The previous sampled value, and the dimensions of the two are... same; Let be the adaptive weight at time t, which is dimensionless; Within the range [0,1].

[0046] As a simplified data fusion model output, it is used for over-limit judgment and linkage confirmation.

[0047] The adaptive weights are determined according to equation (5). (5) in: Let be the adaptive weight at time t, which is dimensionless; The preset minimum weight is dimensionless and ranges from 0.3 to 0.5. The preset maximum weight is dimensionless and ranges from 0.7 to 1.0. The rate of change; The threshold for the rate of change. Greater than zero; and Within the interval (0,1) and Less than .

[0048] In full Kalman filter mode, the programmable controller uses multi-sensor data within a historical data window as the observation sequence and adjusts the observation noise covariance based on the short-term standard deviation of each sensor. The state vector is taken... ,in For sensor fusion estimates, The change rate is the estimated value; the observation vector is the normalized measurement value of each sensor; the state transition matrix is ​​a 2x2 unit augmented matrix (e.g., for a uniform change model, the matrix elements are...). The observation noise covariance matrix is ​​a diagonal matrix, and the diagonal elements are proportional to the short-term standard deviation of the corresponding sensor.

[0049] In step S5, the programmable controller calculates the short-term standard deviation of the signal based on the real-time noise level of each sensor, and generates an adaptive alarm threshold based on the short-term standard deviation of the signal. The adaptive alarm threshold is determined according to formula (6); (6) in: Let t be the adaptive alarm threshold for sensor j at time t, with the same dimensions as the original data of sensor j; The fundamental threshold of sensor j, with dimensions of and same; is a proportionality constant, dimensionless, and its value ranges from 2 to 3; Let be the short-term standard deviation of sensor j within the first time window, with the same dimensions as the original data of sensor j.

[0050] Greater than zero; Non-negative; Greater than zero.

[0051] A support tag that replaces a fixed threshold for exceeding limits and is used for spatial cross-validation.

[0052] Base threshold The length of the first time window is determined by safety condition data during the commissioning calibration phase.

[0053] In step S6, the programmable controller performs multi-level linkage control based on the data from the subsequent real-time monitoring phase. The fused output value from step S4 is extracted. (In simplified mode) or sensor fusion estimate (In filter mode) it is used as the current monitoring parameter.

[0054] When the current monitoring parameter of any sensor exceeds the first linkage threshold, only ventilation and early warning will be activated; only when the monitoring parameter continues to exceed the limit, and the smoke data and temperature data simultaneously exceed their respective second linkage thresholds and remain there for a preset time, will fire extinguishing and audible and visual alarms be triggered.

[0055] The first and second linkage thresholds are dynamically determined based on the adaptive alarm threshold generated in step S5: the first linkage threshold is equal to the corresponding adaptive alarm threshold. The second linkage threshold is an adaptive alarm threshold. Multiply by a preset upgrade factor, which is greater than 1. The preset duration is set according to the hazardous gas diffusion model and is not less than the sum of the sampling periods for smoke data and temperature data.

[0056] In step S7, the programmable controller combines the data from adjacent explosion-proof wireless transceivers in the monitoring matrix to perform spatial cross-verification of the temporary emergency linkage command. The spatial cross-verification result is determined according to equation (7).

[0057] (7) in: The spatial cross-validation result of the event corresponding to the i-th node at time t is dimensionless; Indexes are for adjacent nodes and are dimensionless. Let i be the set of neighboring nodes of node i; For node i and node The spatial weight between the nodes is dimensionless and determined based on the normalized value of the reciprocal of the distance between the two nodes. Adjacent nodes The support marker at time t is dimensionless, and when the node... The value is set to 1 when the current monitoring parameter reaches the corresponding adaptive alarm threshold, and 0 otherwise. The adjacent node support tags are summed according to spatial weights and then divided by the total weights.

[0058] The sum of the weights in the denominator is greater than zero; Not empty.

[0059] The decision on whether to revoke the temporary emergency response order is made by comparing it with the spatial confirmation threshold.

[0060] When spatial cross-validation results When the spatial verification threshold is below the threshold and no danger has been identified in the fine-grained fusion task, the programmable controller executes a reversal command; when the spatial cross-validation result... If the spatial confirmation threshold is not lowered or a hazard is confirmed by a refined fusion task, the linkage strategy should be maintained or upgraded. The spatial confirmation threshold is determined based on the number of adjacent nodes and safety condition calibration data, and its value ranges from 0.3 to 0.7.

[0061] During operation, the commissioning calibration parameters are grouped into versioned parameter groups based on node model, sensor type, measurement unit, and monitoring area conditions. Parameter groups are bound when an event identifier is created, and the parameter version is not switched before the event is retained, revoked, upgraded, or terminated. Parameter updates are written to the backup storage area only when the event is idle and the continuous sampling window meets quality control conditions. After passing integrity verification, the updated parameters are used for subsequent events. If an update fails, the previous valid parameter group is used, the upgrade path for the new event is frozen, and the current time-limited reversible linkage is maintained.

[0062] The rapid threshold, prediction threshold, first linkage threshold, second linkage threshold, and spatial confirmation threshold are all determined by the effective temperature, combustible gas concentration, and smoke concentration data within the safety condition calibration window. Parameter values ​​are written into the risk membership mapping table, threshold calibration table, and linkage strategy table, respectively, and are actually read by the rapid prediction, refined review, spatial cross-validation, and linkage state machine. If any parameter is missing, expired, has inconsistent units, or is version mismatched, the generation of new upgrade or reversal commands stops, the current reversible linkage is maintained, and the system enters a pending review state.

[0063] The rate of change of monitoring parameters, dynamic rate of change threshold, adaptive weight, basic threshold, proportional coefficient, and short-term standard deviation are determined for each sampling period according to equations (2) to (6). The obtained rate of change is written into the fusion mode selection field, the adaptive weight is written into the simplified fusion calculation field, the adaptive alarm threshold is written into the over-limit judgment field of each sensor, and the spatial weight and support flag are written into the weighted summation field of equation (7). When the rate of change or the statistical window is unavailable, the previous valid value is frozen and linkage upgrade is prohibited. Update is resumed after continuous valid sampling reaches the recovery window.

[0064] The control chain for the same event uses a unified time base: ordinary data frames, independent interruption events, refined verification data, spatial cross-validation results, temporary emergency linkage command objects, and execution feedback all carry event identifiers, node identifiers, sampling or sending timestamps, and commissioning calibration versions. The action type, action amplitude or status, linkage execution end, action direction or execution endpoint, sending time, and effective duration in the command object are output by the policy table and consumed by the corresponding execution end; the execution end feedback is written with the same event identifier and command sequence number. The controller only allows maintenance or escalation after matching each field; if the match fails, no subsequent automatic command is generated.

[0065] In this embodiment, the following parameter chain forms a closed relationship: safety operating condition data and sensor configuration determine threshold and membership degree mapping, the mapping output enters the fast prediction and strategy table lookup; the output of equation (2) and equation (3) enters the mode switching, the output of equation (5) enters the fusion value obtained by equation (4), the output of equation (6) enters the over-limit judgment and adjacent node support mark, and the output of equation (7) enters the cancellation, maintenance or upgrade state; each state is driven by the command field with the same event identifier to the corresponding execution end, and is confirmed, terminated or entered the conservative pending review state by the execution feedback.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent linkage control of multiple explosion-proof sensors, applied to a system comprising multiple explosion-proof wireless data transceivers and a programmable controller, wherein each of the explosion-proof wireless data transceivers is configured with a corresponding set of adjacent nodes, characterized in that, include: The explosion-proof wireless data transceiver collects temperature data, combustible gas concentration data, and smoke concentration data, and calculates a rapid prediction value based on the temperature data, combustible gas concentration data, and smoke concentration data. When any of the temperature data, the combustible gas concentration data, or the smoke concentration data exceeds the corresponding rapid threshold, or when the rapid prediction value is not lower than the prediction threshold, the explosion-proof wireless data transceiver first sends a temporary emergency linkage command with an event identifier to the programmable controller through an independent interrupt channel, and periodically uploads the temperature data, the combustible gas concentration data, and the smoke concentration data through a normal data channel. After receiving the temporary emergency linkage command, the programmable controller extracts the temperature data, combustible gas concentration data and smoke concentration data received from the normal data channel in the most recent complete sampling period according to the event identifier, and calculates the refined fusion value; The programmable controller compares the refined fusion value with preset confirmation and cancellation thresholds to perform actions such as maintaining, upgrading, or canceling the temporary emergency linkage command.

2. The explosion-proof multi-sensor intelligent linkage control method according to claim 1, characterized in that, The calculation of the rapid prediction value based on the temperature data, the combustible gas concentration data, and the smoke concentration data includes: The temperature data, the combustible gas concentration data, and the smoke concentration data are respectively subjected to minimum-maximum normalization. The membership degree of the normalized temperature data, combustible gas concentration data, and smoke concentration data is calculated and output using a piecewise linear membership function. The maximum value of the three membership degree outputs is taken as the fast prediction value; the calculation formula for the fast prediction value is as follows: in: Let be the rapid prediction value of the i-th explosion-proof wireless data transceiver at time t, which is dimensionless; For temperature data, This is data on combustible gas concentration. Data on smoke concentration; The temperature membership function Let f be the membership function of combustible gas concentration. Let f be the membership function of smoke concentration. ( ), ( ), ( ) are the piecewise linear membership functions output after normalization of the corresponding sensor data.

3. The explosion-proof multi-sensor intelligent linkage control method according to claim 1, characterized in that, The calculation of the refined fusion value includes calculating the refined fusion value using a weighted summation method; The programmable controller compares the refined fusion value with preset confirmation and cancellation thresholds respectively to perform actions such as maintaining, upgrading, or canceling the temporary emergency linkage command, including: When the refined fusion value is not lower than the confirmation threshold, a danger is confirmed, and the temporary emergency linkage instruction is maintained or upgraded. When the refined fusion value is lower than the revocation threshold, no danger is confirmed, and a revocation command is sent to the corresponding explosion-proof wireless data transceiver and linkage execution terminal. When the refined fusion value is between the cancellation threshold and the confirmation threshold, the system maintains the linkage state of the previous moment and triggers the continuous refined fusion task of the next cycle.

4. The explosion-proof multi-sensor intelligent linkage control method according to claim 3, characterized in that, The method further includes steps for detecting sudden changes in operating conditions and switching modes, specifically: The programmable controller calculates the rate of change of the monitoring parameters based on the temperature data, the combustible gas concentration data, and the smoke concentration data in adjacent sampling periods; The rate of change of the monitored parameters is compared with the dynamic rate of change threshold. When the rate of change of the monitored parameter exceeds the dynamic rate of change threshold, the simplified data fusion mode is switched and the fusion output value is calculated as the current monitored parameter. At the same time, a high sampling frequency status command is sent to the corresponding explosion-proof wireless data transceiver. When the simplified data fusion mode is in effect, and the rate of change of the monitored parameter is continuously lower than the preset recovery threshold and maintains a preset recovery time, the sensor fusion estimate is calculated using the full Kalman filter mode as the current monitored parameter; the recovery threshold is lower than the dynamic rate of change threshold.

5. The explosion-proof multi-sensor intelligent linkage control method according to claim 4, characterized in that, in: Let be the rate of change of the monitoring parameters of the i-th node at time t. The temperature data is at time t. This is the temperature data from the previous sampling time. This is the preset range reference value for temperature data. Here is the data on the concentration of combustible gas at time t. This is the combustible gas concentration data from the previous sampling time. This is the preset range reference value for combustible gas concentration. Here is the smoke concentration data at time t. This is the smoke concentration data from the previous sampling time. This is the preset range reference value for smoke concentration. The sampling interval is... The formula for calculating the dynamic change rate threshold is as follows: in: The threshold for the rate of change at time t; The average rate of change within the historical rate of change window; The standard deviation of the rate of change within the historical rate of change window; This is the proportionality coefficient.

6. The explosion-proof multi-sensor intelligent linkage control method according to claim 4, characterized in that, The step of switching to simplified data fusion mode and calculating the fusion output value as the current monitoring parameter includes: Only the data from the two most recent sampling periods are retained, and an adaptive weighted average of the rate of change is performed on each sensor to obtain the fused output value. The calculation formula is as follows: in: Let be the fused output value of the i-th node at time t. This is the current sampled value. The previous sampled value, Let be the adaptive weights at time t. Within the range [0,1]; As a simplified data fusion model output, it is used for over-limit judgment and linkage confirmation; The formula for calculating the adaptive weights is as follows: in: Let be the adaptive weights at time t. To preset the minimum weight, To preset the maximum weight, The rate of change; The threshold for the rate of change. Greater than zero; and Within the interval (0,1) and Less than .

7. The explosion-proof multi-sensor intelligent linkage control method according to claim 4, characterized in that, The method further includes an adaptive threshold generation step based on the real-time noise level, specifically: The programmable controller calculates the short-term standard deviation of the historical data of each sensor within the first time window. Based on the corresponding base threshold and the short-term standard deviation, adaptive alarm thresholds are generated for each sensor; the calculation formula for the adaptive alarm threshold is as follows: in: Let be the adaptive alarm threshold for sensor j at time t. The basic threshold of sensor j, This is the proportionality coefficient. Let j be the short-term standard deviation of sensor j within the first time window. Greater than zero; Non-negative; Greater than zero; It replaces fixed thresholds for exceeding limits and serves as a support marker for spatial cross-validation; The programmable controller uses the short-term standard deviation of each sensor as the observation noise covariance and inputs it into the complete Kalman filter mode for state updates.

8. The explosion-proof multi-sensor intelligent linkage control method according to claim 7, characterized in that, The method further includes multi-level linkage execution control steps, specifically: The programmable controller compares the current monitoring parameters of each sensor with the first linkage threshold and the second linkage threshold, wherein the first linkage threshold is equal to the corresponding adaptive alarm threshold, and the second linkage threshold is the adaptive alarm threshold multiplied by a preset escalation coefficient. When the current monitoring parameter of any sensor exceeds the first linkage threshold, ventilation and early warning operations are initiated. When the current monitoring parameter continues to exceed the limit, and the current monitoring parameter corresponding to the smoke concentration data and the current monitoring parameter corresponding to the temperature data both exceed their respective second linkage thresholds and remain so for a preset time, the fire extinguishing and audible and visual alarm operations are triggered.

9. The explosion-proof multi-sensor intelligent linkage control method according to claim 7, characterized in that, The method also includes a spatial cross-validation step for temporary emergency response commands, specifically: The programmable controller obtains the current monitoring parameters of each neighboring node in the neighboring node set at the corresponding time. When the current monitoring parameter of a certain neighboring node reaches the corresponding adaptive alarm threshold, the support flag of the neighboring node is set to 1, otherwise it is set to 0. The spatial cross-validation result is calculated by combining the spatial weights between the explosion-proof wireless data transceiver and each adjacent node, as well as the corresponding support tags. The formula for calculating the spatial cross-validation result is as follows: in: Let be the spatial cross-validation result of the event corresponding to the i-th node at time t. Indexing adjacent nodes Let i be the set of neighboring nodes of node i; For node i and node Spatial weights between them Adjacent nodes Support tag at time t, when node The value is set to 1 when the current monitoring parameter reaches the corresponding adaptive alarm threshold, and 0 otherwise. When the spatial cross-validation result is lower than the spatial confirmation threshold and the refined fusion value is not confirmed as dangerous, the programmable controller sends the revocation command to the corresponding explosion-proof wireless data transceiver and linkage execution terminal.

10. The explosion-proof multi-sensor intelligent linkage control method according to claim 1, characterized in that, The method further includes a control chain time base binding step, specifically: Establish a set of commissioning calibration parameters with version numbers according to node model, sensor type, measurement unit, and monitoring area operating conditions; When generating the temporary emergency linkage command and binding the event identifier, the current version of the commissioning calibration parameter group is locked; the version of the commissioning calibration parameter group is not switched before the event is retained, revoked, upgraded or terminated.

Images