Point inspection method and system for dew point sensor

Abstract

The application relates to the technical field of sensors, and discloses a point inspection method and system for a dew point sensor. The method comprises the following steps: initially, all probes are sealed and closed, one probe is selected and opened, and on-site dew point routine detection is performed; when a point inspection period is reached, a special point inspection probe is opened, and dew point data of the routine detection probe and the point inspection probe are synchronously collected; difference operation is performed on the two groups of data, the result is compared with a preset threshold value to obtain a probe working state; if the result exceeds the threshold value, a standby probe is enabled, and the abnormal probe and the point inspection probe are restored to be sealed and closed; the point inspection is circularly performed, the point inspection probe is compared with the in-use probe, and a fault prompt is output when all reusable probes are abnormal. The application does not need manual intervention, can realize online self-inspection, state determination and fault early warning of the probe, solves the problems of single-probe dependence on manual calibration and data distortion, and improves detection reliability and process control precision.

Description

Technical Field

[0001] This application relates to the field of sensor technology, and more specifically, to a method and system for inspecting a dew point sensor. Background Technology

[0002] Dew point sensors are widely used for humidity and dew point temperature detection in industrial settings. Capacitive dew point sensors, with a moisture-sensitive thin-film capacitor as the core sensing device, offer advantages such as fast response and high accuracy. Existing capacitive dew point sensors mostly employ a single-probe structure combined with periodic calibration strategies for condition assessment, i.e., performing offline calibration at fixed intervals to correct detection deviations. However, these sensors are mostly used online in real time, and probe failure time is affected by multiple factors such as environmental corrosion, media contamination, and thin-film aging, exhibiting non-linear decay characteristics. Periodic calibration strategies rely on manual intervention, making it impossible to predict failure points. The risk of data distortion accumulates continuously within the calibration interval, seriously affecting industrial production safety and the accurate control of process parameters. Summary of the Invention

[0003] The main purpose of this application is to provide a method and system for inspecting dew point sensors, which aims to solve the technical problems of existing capacitive dew point sensors that rely on manual labor for periodic offline calibration with a single probe, cannot predict failure points, and cause data distortion during calibration intervals, affecting production safety and process control.

[0004] The first aspect of this application proposes a method for inspecting a dew point sensor, comprising: Initially, all probes are kept sealed and closed. Then, one probe is selected and turned on to continuously perform routine on-site dew point testing. When the preset inspection cycle is reached, the dedicated inspection probe is turned on, and dew point detection data corresponding to the conventional detection probe and the inspection probe are collected simultaneously. The difference between the two sets of dew point detection data is calculated, and the calculation result is compared with the preset judgment threshold to obtain the probe working status judgment result. If the comparison result exceeds the threshold, the backup probe will be activated to replace the regular dew point detection, and the original abnormal detection probe and the inspection probe will be restored to the sealed and closed state. The inspection cycle is triggered in a round-by-round manner. Each time, the data is compared with the currently used detection probe through the inspection probe. When all reusable detection probes are abnormal, a fault prompt signal is output.

[0005] Furthermore, in the initial state, all probes are kept sealed and closed, and one probe is selected to be turned on and continuously perform routine on-site dew point testing, including: Read all probe intrinsic parameters, baseline data and five dynamic parameters, construct a multi-dimensional information mapping table, and uniformly mark all probes as sealed and disable the acquisition channel; Based on the constructed information mapping table, a hierarchical screening is carried out. The priority index is calculated by combining hardware qualifications, calibration accuracy and dynamic parameters, and a unique initial candidate detection probe is selected from all probes. Check the sealing status of the candidate probes. If they are qualified, issue an opening command and update the status. If they are not qualified, remove them and re-screen them until a compliant probe is selected and the selection result cannot be changed. Assign independent storage and sampling permissions to the selected probe, activate the dedicated acquisition channel and block the access permissions of other probes, and control it to continuously carry out on-site dew point detection and upload detection data.

[0006] Furthermore, the step of activating the dedicated inspection probe and simultaneously collecting dew point detection data corresponding to the conventional detection probe and the inspection probe when the preset inspection cycle is reached includes: The background runs a timing task and matches the built-in inspection cycle parameters. The cycle is determined by the system timestamp. Once the cycle conditions are met, an inspection trigger event is generated, the dedicated inspection probe is turned on, and the preheating process begins. After the preheating is completed, the sampling sequence numbers of the two sets of probes are reset and cleared to zero. The inspection probes are unlocked, initialized, handshaked, and self-checked to maintain the uninterrupted online acquisition link of the regular detection probes. Send a synchronous trigger enable signal, define the acquisition time window and allocate independent communication time slots, issue a unified sampling scheduling command, control the two probes to carry out sampling in parallel under the same sequence and working conditions and mark the sampling sequence number; Based on the sampling sequence number, two sets of data are paired and encapsulated point-to-point, collected into a dedicated data buffer for inspection, and aligned according to the queue rules. The normalized data is then stored in the corresponding comparison partitions and the regular probe data queue status is restored.

[0007] Further, the step of performing a difference calculation on the two sets of dew point detection data, comparing the calculation result with a preset judgment threshold, and obtaining the probe working status judgment result includes: Read two sets of dew point detection data with time alignment, complete precise pairing based on synchronization identifier, remove abnormal and invalid data points and define judgment data window, construct standard comparison dataset, and generate signed difference sequence by point-by-point operation; The dataset is segmented and divided into sliding windows. The deviation statistics are calculated segment by segment. The proportion of outliers and the distribution of the sign of the difference are statistically analyzed. The cumulative sum sequence is calculated recursively, and the out-of-bounds points are statistically analyzed. The system retrieves various preset constraint thresholds and judgment limits to conduct comprehensive cross-comparison and verification of segmented deviation, overall deviation, data fluctuation, sign consistency, and cumulative and out-of-bounds situations. Based on the combination and matching relationship of multi-dimensional verification indicators, the normal state of the probe is distinguished from different abnormal types, and a fixed probe working status judgment result is generated.

[0008] Furthermore, the steps of segmenting the dataset and dividing it into sliding windows, calculating the deviation statistics segment by segment, statistically analyzing the proportion of outliers and the distribution of the difference sign within the window, recursively calculating the cumulative sum sequence, and completing the statistics of out-of-bounds points include: Read the signed difference sequence and initialize the cumulative sum benchmark, allocate a dedicated storage area for the sequence, perform validity checks on the difference data, remove invalid data, and complete the classification labeling according to positive and negative types; The cumulative operation is performed separately according to positive and negative categories. The cumulative results of the two types of differences are recorded independently. Simultaneously, the first-level cumulative sum corresponding to the original difference and the second-level cumulative sum sequence corresponding to the absolute value of the difference are generated recursively. The difference data is divided into segments as a whole, and an independent cumulative sum sequence is established for each segment. The deviation statistics are calculated for each segment, the outlier points in each segment are counted and the proportion of local outliers is calculated to form a segmented cumulative sequence. The system iterates through the cumulative sums and segmented subsequences at each level, compares and screens outbound points and abnormal segments, summarizes various statistical indicators, and stores the statistical results and each cumulative sum sequence into their respective independent storage areas.

[0009] Furthermore, the step of activating a backup probe to replace the regular dew point detection if the comparison result exceeds the threshold, and restoring the original anomaly detection probe and the inspection probe to a sealed and closed state, includes: After determining that the threshold comparison result is abnormal, lock the two types of probe device identifiers and establish a dedicated control link to isolate external signal interference. After reading the status flag bit, issue shutdown commands in sequence and receive corresponding feedback frames. Verify the timing information of the feedback frame to confirm that the instruction is executed effectively, issue a collection termination instruction and shut down the collection channel and communication port, clear the collection buffer, and reissue the compliant shutdown instruction after verifying the communication link through test frame interaction. The system sequentially issues sampling pause, moderate shutdown and complete shutdown control commands in stages, updates the sealing status flag after receiving execution feedback, initiates a bus release request and sets the communication port to the isolated state. The system solidifies the sealing configuration of the two types of probes and completes the status verification. It then disables the relevant control and acquisition interfaces, resets the buffer, address pointer, and counting resources, sends an enable command to the backup probe, and completes the switching and takeover of the routine dew point detection task.

[0010] Furthermore, the steps of locking the identifiers of the two types of probe devices and establishing a dedicated control link after the judgment threshold comparison result is abnormal, isolating external signal interference, reading the status flag bit, issuing shutdown commands sequentially level by level, and receiving corresponding feedback frames include: Lock onto the target probe device identifier, establish a dedicated control link and isolate external signal interference, read the status flag bit and complete the validity verification, and create a command interaction waiting record table; Send the first round of shutdown commands to the two types of probes and configure exclusive serial numbers, open the dedicated receiving channel, start the listening mechanism and activate the receiving interrupt to capture feedback frames; After capturing the feedback frame, stop listening and interrupting, record time information, verify the relevant identifiers and codes of the feedback frame, update the record table and send a completion signal after a match is found. After receiving the completion signal, a second round of shutdown command is issued, the interrupt conditions are reconfigured and the listening is restarted, the feedback frame verification is completed, the execution status is marked and the two-level command interaction is confirmed to be complete.

[0011] Furthermore, the steps of sending the first round of shutdown commands to the two types of probes and configuring exclusive serial numbers, opening dedicated receiving channels, starting the monitoring mechanism, and activating the receiving interrupt to capture feedback frames include: Send the first round of shutdown commands with sequence identifiers to both types of probes, open the dedicated receiving channel, save the task context, and define the blocking waiting time boundary; The original execution flow of the program is blocked, the task suspension and semaphore wait functions are called, and the current task is moved into the blocking queue and the semaphore wait queue. Shield system-irrelevant interrupts and low-priority task scheduling, enable dedicated receive interrupts, release semaphores and wake up blocked processes after matching the target feedback frame; After the process resumes running, it verifies the semaphore status and timing information, clears the task suspension flag, closes the dedicated receive channel, destroys the semaphore, and releases the blocked waiting state.

[0012] Furthermore, the step of triggering the inspection cycle in successive cycles, comparing data between the inspection probe and the currently used detection probe each time, and outputting a fault indication signal when all reusable detection probes malfunction, includes: Construct a reusable probe circular queue and initialize it according to priority. Independently set up dedicated inspection probes, initialize the inspection state machine and bind the probe identifier. After each round of inspection cycle, establish a fixed pairing relationship between the probes in use and the inspection probes. After the periodic trigger, the inspection state machine is switched, and the paired probe is synchronously driven to complete the unified timing acquisition, cache the acquired data and generate an inspection log, recording the inspection-related identifiers and data summary information; Extract the two detection data streams and perform validity verification and format standardization. Construct a standardized dataset. Use the inspection probe data as a benchmark to perform difference calculation and threshold comparison. Switch the state machine and update the inspection log based on the results. When a probe is found to be abnormal, it is removed from the loop queue, and a subsequent probe is used to take over the detection and the relevant probe status is reset. The inspection process is continuously executed in a loop. When the queue of reusable probes is exhausted, a fault prompt signal is output.

[0013] A second aspect of this application also proposes a dew point sensor inspection system, comprising: The initial control module is used to keep all probes sealed and closed in the initial state, select one probe to turn on and continuously perform routine on-site dew point testing; The periodic acquisition module is used to activate the dedicated inspection probe and simultaneously acquire dew point detection data corresponding to the conventional detection probe and the inspection probe when the preset inspection cycle is reached. The calculation and comparison module is used to perform difference calculation on two sets of dew point detection data, compare the calculation result with the preset judgment threshold, and obtain the probe working status judgment result. The abnormal switching module is used to activate the backup probe to replace the regular dew point detection if the comparison result exceeds the threshold, and restore the original abnormal detection probe and the inspection probe to the sealed closed state. The cyclical maintenance module is used to trigger the inspection cycle in rounds. Each time, the data is compared with the currently used detection probe through the inspection probe. When all reusable detection probes are abnormal, a fault prompt signal is output.

[0014] The first aspect of this plan brings the following benefits: This application enables online self-inspection of probes without manual intervention by using multi-probe cyclic inspection and synchronous data comparison. It can determine the working status of probes in real time, predict failure points, and effectively solve the problems of single probe reliance on manual offline calibration and data distortion within calibration intervals. It avoids the impact of abnormal detection data on production safety and process control, realizes full autonomous control of dew point detection, and improves detection reliability and process control accuracy. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a method for inspecting a dew point sensor according to an embodiment of this application. Figure 2 This is a schematic diagram of the structure of a dew point sensor inspection system according to an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a computer device according to an embodiment of this application; The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] Those skilled in the art will understand that, unless explicitly stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in the specification of this application means the presence of features, integers, steps, operations, elements, modules, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components, and / or groups thereof. It should be understood that when an element is “connected” or “coupled” to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein may include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any modules and all combinations of one or more associated listed items.

[0018] Those skilled in the art will understand that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0019] Reference Figure 1 This application provides a method for inspecting a dew point sensor, including: S1: In the initial state, all probes are kept sealed and closed. Select one probe to turn on and continue to perform routine on-site dew point testing. S2: When the preset inspection cycle is reached, the dedicated inspection probe is turned on to simultaneously collect dew point detection data corresponding to the conventional detection probe and the inspection probe. S3: Perform difference calculation on the two sets of dew point detection data, compare the calculation result with the preset judgment threshold, and obtain the probe working status judgment result; S4: If the comparison result exceeds the threshold, the backup probe will be activated to replace the regular dew point detection, and the original abnormal detection probe and the inspection probe will be restored to the sealed and closed state. S5: Triggers the inspection cycle in cycles, and compares the data with the currently used detection probes each time. If all reusable detection probes are abnormal, a fault prompt signal is output.

[0020] In step S1, the probes are pre-classified into types, including reusable detection probe groups, dedicated benchmark inspection probes, and spare probes. Dedicated benchmark inspection probes are normally kept sealed and closed, only opened briefly during inspection cycles, and do not participate in long-term routine testing. The initial state benchmark is determined by uniformly setting all dew point probes in the field to a sealed and closed state. A total of 8 probes are deployed in the field, numbered sequentially from P01 to P08. Each probe has unique factory calibration parameters, static baseline data, and five dynamic parameter identifiers. Next, the input parameters for initial screening and state configuration are prepared: the probe basic information set includes the hardware numbers of the 8 probes, factory calibration error, hardware batch, and static baseline deviation; the dynamic parameter set includes the historical service time, cumulative number of tests, environmental exposure risk score, fault interval duration, and power supply voltage margin for each probe; the sealed state identifier register occupies 1 storage bit per probe, with 0 representing sealed and 1 representing open; there are 8 probe acquisition channels, with channel numbers corresponding one-to-one with hardware numbers. The system then powers on and completes initialization, reads all inherent and dynamic parameters from the eight probes, constructs a multi-dimensional probe basic information mapping table, sets the sealing status flag of all probes to 0, and disables all eight acquisition channels and sampling processes. Following a layer-by-layer screening process based on hardware batch, repair records, calibration accuracy, static baseline, and dynamic parameter weighting rules, a comprehensive priority index is calculated, selecting probe P03 as the sole initial candidate detection probe. After verifying that P03's sealing status flag is 0 and there is no open / occupied marker, an open command 0x01 is issued, and a valve feedback status code 0xAA is received. The P03 sealing status flag is then updated to 1, and the open timestamp is recorded. An independent fixed data storage partition and dedicated sampling scheduling permissions are allocated to P03, and the P03 data acquisition path is activated separately, setting a 1-second single-time continuous acquisition mode. The remaining probes (P01, P02, P04~P08) maintain a sealing status flag of 0 and their channels remain disabled, blocking the regular detection access permissions of the other probes. P03 continuously acquires field dew point data and uploads it to the main control unit via the Modbus protocol. In this system, a sealing status indicator value of 0 represents the probe being sealed closed, and 1 represents the probe being open. A smaller comprehensive priority index value indicates higher probe compatibility, and a 1-second acquisition cycle represents a fixed-sequence sampling interval. Step S1 transforms the disordered initial state of multiple probes into a quantifiable operational benchmark for fixed-point routine testing of a single probe through unified management of the entire probe status, parameter table creation, hierarchical screening, and channel configuration. This establishes a stable operating condition foundation for subsequent point inspection acquisition in S2 and difference comparison in S3.

[0021] In step S2, the preset inspection cycle benchmark is first determined to be the system default inspection cycle of 25,920,000 seconds. The conventional detection probe P03 selected in S1 is used, and the dedicated inspection probe is fixed as P09. Next, the input parameters for synchronous acquisition are prepared: the system timing register stores the preset inspection cycle of 25,920,000 seconds; the dual-probe sampling frequency is fixed at 1 second / time, with 100 sampling points collected continuously in a single sampling; the communication command code includes an activation command 0x02 and a sampling reset command 0x04; the synchronous trigger pulse is high for 10 milliseconds, and the inspection probe warm-up time is fixed at 60 seconds. Then, an independent timing task runs in the system background, comparing the system timestamp in real time. When the difference between the current timestamp and the previous inspection timestamp reaches 25,920,000 seconds, the inspection flag is triggered. An activation command 0x02 is sent to the dedicated inspection probe P09, and after receiving the valve feedback status code 0xBB, the 60-second warm-up phase begins. After preheating, a sampling sequence number reset command 0x04 is simultaneously sent to both the regular detection probe P03 and the inspection probe P09, clearing the sampling sequence number counters of both probes and unifying the sampling start sequence number to 0. The system sends a hardware trigger pulse enable signal, which maintains a high level for 10 milliseconds before automatically pulling low. The synchronous trigger module simultaneously sends a sampling start pulse to P03 and P09. A fixed acquisition time window is defined, and a dedicated communication time slot is allocated. The two probes acquire 100 sampling points in parallel under the same sequence and operating conditions at a sampling frequency of 1 second per sampling. The two sets of dew point data are paired point by point according to the sampling sequence number, encapsulated with a unified window identifier and timing number, and stored in the inspection-dedicated data buffer. After timing alignment, the data is stored in the regular area and the inspection area respectively, and the P03 regular acquisition queue is restored to normal writing. The period value of 25,920,000 seconds corresponds to the fixed inspection interval, the 100 sampling points are the single comparison data capacity, the command codes 0x02 and 0x04 represent fixed function control codes, and the 10-millisecond pulse is the synchronization timing reference. Step S2 completes high-precision time-series synchronous acquisition of P03 and P09 through periodic determination, probe preheating, sequence number clearing, and hardware synchronization triggering. In each round of inspection, the baseline self-verification of the dedicated reference inspection probe is performed first. If the drift of the dedicated reference inspection probe exceeds the preset threshold, the current round of inspection is terminated and the reference fault is reported. The in-use probe judgment is not carried out. This provides standardized raw data with consistent timing and the same operating conditions for step S3 to carry out dew point data difference calculation and threshold comparison.

[0022] In step S3, the comparison benchmark is first determined as 100 sets of time-aligned dew point data from the conventional detection probe P03 and the spot inspection probe P09, which were synchronously acquired in S2. The preset judgment threshold is fixed at ±2℃. Next, the input parameters for difference calculation and threshold comparison are prepared: the P03 measurement point sequence contains 100 dew point values, and the P09 measurement point sequence contains 100 dew point values; the preset deviation threshold is ±2℃, the sliding window length is 20 points, and the step size is 10 points; the first-level cumulative sum threshold is ±10℃, the second-level cumulative sum threshold is 0~15℃, and the segmentation rule is that every 25 data points constitute one segment. Then, the 100 sets of normalized valid dew point data from P03 and P09 are read from the synchronous comparison buffer, and point-to-point matching is performed sequentially, eliminating invalid data points with frame anomalies or incorrect markings. The signed difference between the two sets of data is calculated point by point, generating a 100-point difference sequence. The system divides local windows into 20-point sliding windows with 10-point increments, counting the number of outliers within each window whose absolute difference exceeds the ±2℃ threshold. The 100 data points are divided into four independent segments of 25 points each, and the median deviation and overall average deviation of each segment are calculated. First-level signed cumulative sum sequences and second-level absolute value cumulative sum sequences are generated recursively. The number of outliers exceeding ±10℃ in the first-level cumulative sum and the number exceeding 15℃ in the second-level cumulative sum are counted. The proportion of positive and negative points in the difference sequence is calculated to obtain the sign consistency ratio. The segment median, overall deviation, window anomaly ratio, and number of outliers in the cumulative sum are simultaneously cross-checked with each preset threshold. If multiple indicators are within the threshold range, P03 is considered to be working normally, and status code 0x00 is output; if the outlier window, sign ratio, and number of outliers all exceed the limits, it is considered a systematic cumulative drift anomaly, and status code 0x09 is output; other cases correspond to random local anomalies or complex anomalies, and the corresponding status code is output. ±2℃ is the maximum allowable deviation threshold for dew point, 100 points is the total amount of data for a single judgment, the cumulative sum threshold is used to quantify the degree of long-term offset, and the status code represents the unique fixed probe working type. If communication timeout, sampling data packet loss, or timing alignment disorder occurs, it will automatically retry 1 to 2 times. If the retry is still abnormal, the current round of inspection is marked as invalid and probe switching is not triggered. Step S3 calculates the point-by-point difference of the dew point data from the two probes, performs segmented statistics, verifies the cumulative sum, and cross-compares it with the threshold to output a standardized probe status judgment result, providing a quantitative judgment basis for step S4 of abnormal probe shutdown and backup probe switching.

[0023] In step S4, the abnormal object is first identified as the conventional detection probe P03, whose performance exceeded the threshold as determined in S3; the inspection probe is P09; and the backup probe to be activated is P04, the first probe in the circular queue. The preset instruction feedback timeout threshold is 500 milliseconds, and the shutdown instruction codes are 0x03 and 0x0E, with corresponding feedback frames of 0xCC and 0xEE. Next, the input parameters for probe switching and sealing closure are prepared: the probe device identifier register stores the P03 and P09 numbers; the status flag bit is 0 for sealing closure and 1 for opening; the three-stage control durations are 2 seconds for sampling pause and 3 seconds for valve partial closure; the bus release instruction code is 0x18, and the solidification configuration instruction code is 0x16. Then, the system determines that the P03 comparison result exceeds the ±2℃ preset threshold, immediately locks the P03 and P09 device identifiers, and establishes a dedicated control link to isolate bus interference. After reading that both probe status flags are 1, the first shutdown command 0x03 and the second shutdown command 0x0E are issued sequentially. The system waits for corresponding feedback frames 0xCC and 0xEE, verifying that the difference between the two feedback timestamps is less than 100 milliseconds to confirm the command execution is valid. An error check and test frame is issued. After verifying that the communication module responds normally, a three-stage shutdown process is executed. First, a sampling pause command is issued and maintained for 2 seconds. Then, a valve half-close command is issued, reducing the valve opening to 50% and maintaining it for 3 seconds. Finally, a complete shutdown command is issued, setting the valve opening to zero. After receiving the valve feedback frame, the status flags of both probes are updated to 0, a bus release request 0x18 is initiated, and the communication port switches to high-impedance isolation mode. A configuration solidification command 0x16 is issued, writing the sealing status to non-volatile storage, disabling the P03 and P09 acquisition and control interfaces, clearing the data buffer, and resetting the sampling counter. An opening command 0x01 is issued to the backup probe P04. After receiving feedback 0xAA, the status flag is updated to 1, and P04 takes over the routine dew point detection in the field. The 500-millisecond timeout period is for command timeout determination, while the 2-second and 3-second periods are fixed durations for valve stage control. Status flags 0 and 1 correspond to two fixed probe operating states, and each hexadecimal command code corresponds to a single control function. Step S4, through abnormal probe locking, tiered command interaction, phased valve closure, status solidification, and backup probe activation, completes the abnormal probe seal reset and seamless handover of the detection task, providing a compliant probe status basis for the S5 cyclical inspection and rotation.

[0024] In step S5, the reusable probe cyclic queue is first determined to consist of P01, P02, P04, P05, P06, P07, and P08. The dedicated inspection probe remains P09. After switching in S4, the currently used inspection probe is P04, and the inspection state machine is initially in an idle state. Next, the input parameters for the cyclic inspection are prepared: the inspection cycle is fixed at 25,920,000 seconds; the state machine includes idle, ready, acquisition, comparison, judgment, and switching states; the probe hash calculation rule is the accumulation of identifier bytes and modulo 256; the reusable probe queue capacity is 7 points, and the fault trigger condition is a queue length of 0. The system initializes the reusable probe cyclic queue, enqueuing P01, P02, P04, P05, P06, P07, and P08 sequentially according to priority, setting the queue head P04 as the currently used inspection probe. The device identifiers of P04 and P09 are hashed separately to establish a one-to-one inspection pairing mapping relationship. After each inspection cycle, the inspection state machine switches from idle to ready state, reads the identifier of the head probe in the queue, and enters the acquisition state after confirming a match. Simultaneously, P04 and P09 are triggered to sequentially acquire 100 sets of dew point data, generating an inspection log that records the timestamp, probe number, and sampling data summary. Upon entering the comparison state, the two sets of data are time-aligned, difference calculated, and threshold compared, using the ±2℃ judgment threshold. When a probe is determined to be normal, the state machine returns to idle state, the head probe position remains unchanged, and the timestamp for the next inspection is updated. When a probe is determined to be abnormal, the head probe is removed from the circular queue, the queue length is decreased by 1, the next probe in the queue becomes the new head probe and switches to normal detection mode, and the original abnormal probe returns to a sealed closed state. The inspection process is continuously executed in cycles, sequentially completing the data comparison between the in-use probes and P09, status determination, and probe rotation. When all probes in the reusable probe circular queue are determined to be abnormal and the queue length is reduced to 0, the system outputs a fixed fault indication signal. The queue length represents the number of remaining available probes, the hash value is a unique index for probe pairing, and the six states of the state machine strictly limit the inspection process flow. A queue length of 0 indicates that there are no available backup probes, and temporarily abnormal probes are added to the fault reset queue. Probes that meet the self-calibration conditions can be re-enqueued for reuse. A fault indication signal is output only when all reusable probes and backup probes are determined to be permanently faulty. Step S5 achieves automatic inspection and maintenance throughout the entire lifecycle through reusable probe cyclic queue scheduling, state machine step-by-step flow, round-by-round inspection comparison, and probe rotation. A fault signal is output in a timely manner after all reusable probes fail, completing the closed loop of the entire dew point sensor inspection process.

[0025] In one embodiment, the initial state involves keeping all probes sealed and closed, selecting one probe to be turned on, and continuously performing routine on-site dew point testing, including: S10: Read all probe intrinsic parameters, baseline data and five dynamic parameters, construct a multi-dimensional information mapping table, uniformly mark all probes as sealed and close and disable the acquisition channel; S11: Based on the constructed information mapping table, hierarchical screening is carried out. The priority index is calculated by combining hardware qualifications, calibration accuracy and dynamic parameters, and a unique initial candidate detection probe is selected from all probes. S12: Check the sealing status of the candidate probe. If it is qualified, issue an opening command and update the status. If it is not qualified, remove it and re-screen until a compliant probe is selected and the selection result cannot be changed. S13: Assign independent storage and sampling permissions to the selected probe, activate the dedicated acquisition path and block the access permissions of other probes, and control it to continuously carry out on-site dew point detection and upload detection data.

[0026] In this embodiment, the first step is to read all probe inherent parameters, baseline data, and five dynamic parameters to construct a multi-dimensional information mapping table. All probes are then uniformly marked as sealed and closed, with their acquisition channels disabled. The conventional probes P01~P08 and the dedicated inspection probe P09 deployed on-site are used. All read data is the original, fixed data after system power-on initialization, maintaining consistency with subsequent screening process data. Probe inherent parameters include hardware serial number, factory calibration error, hardware batch, and factory hardware qualification. Baseline data is the static baseline deviation benchmark value for each probe. The five dynamic parameters include historical service time, cumulative number of tests, environmental exposure risk score, fault interval duration, and power supply voltage margin. Taking P03 as an example, its factory calibration error is 0.3℃, hardware batch code is 202503, static baseline deviation is 0.08℃, historical service time is 1200 hours, cumulative number of tests is 860,000, environmental exposure risk score is 15 points, fault interval is 360 days, and power supply voltage margin is 0.4V. The system records all information from the nine probes one by one, forming a standardized multi-dimensional information mapping table. It sets the sealing status flag of all probes to 0, indicating that the seal is closed, and simultaneously disables all probe acquisition channels and sampling processes, locking the initial working state of the hardware, thus providing a complete data foundation for stratified screening.

[0027] Secondly, a tiered screening process is conducted based on an information mapping table. A priority index is calculated by combining hardware qualifications, calibration accuracy, and dynamic parameters to select the unique initial candidate test probe. Hardware qualifications are screened based on batch age and the presence or absence of repair records. Calibration accuracy is screened based on factory calibration error and baseline correction coefficient. Dynamic parameters are used to calculate a comprehensive priority index with fixed weights: service life (25%), number of tests (20%), environmental risk (25%), failure interval (15%), and voltage margin (15%). The system first eliminates overdue batches and probes with repair records, then retains probes with calibration errors between 0 and 1.5℃ and baseline correction coefficients between 0.9 and 1.1. The comprehensive priority index is calculated for each retained probe. P03's index is 18.5, the minimum among all candidate probes, and it is selected as the unique initial candidate test probe. If the indices are the same, the shortest historical service life is used as the criterion, and the selected result is fixed and cannot be changed.

[0028] Next, the sealing status indicator of the candidate probes is checked. If it passes, an opening command is issued and the status is updated; otherwise, it is removed and re-selected. The system reads the P03 sealing status indicator bit, where 0 indicates a compliant closed state and 1 indicates an open state. In this case, the P03 indicator is 0, meeting the activation conditions. The system issues an opening command 0x01 to P03. After the probe returns a status code 0xAA, the system updates its status indicator to 1 and records the opening timestamp. If the candidate probe's indicator is 1, it is directly removed. The stratified screening and status verification process is repeated among the remaining probes until a compliant probe with an indicator of 0 is selected. Once the selection result is confirmed, it will not be changed. Under normal operating conditions, the selection is locked and effective. If the selected probe experiences a sudden communication failure or self-test failure, the stratified screening and selection process is restarted for reselection. Then, independent storage and sampling permissions are assigned to the selected probe, its dedicated acquisition channel is activated, access permissions for other probes are blocked, and dew point detection is continuously performed and data is uploaded. The system allocates an independent data storage partition for P03, configures dedicated sampling scheduling permissions, activates the acquisition path and sampling process separately, sets the sampling period to 1 second, and maintains uninterrupted, continuous acquisition. Simultaneously, routine detection access permissions for P01, P02, P04 through P09 are blocked, maintaining a sealed, closed, and disabled state. P03 acquires field dew point data once per second, encapsulates it via the Modbus protocol, and uploads it to the main control unit, continuously undertaking routine field dew point detection tasks. This step establishes a multi-dimensional parameter ledger for the probes, selects optimal probes through hierarchical screening and priority indices, strictly controls sealing compliance, and implements isolated permission configurations to achieve standardized and automatic initial probe deployment, ensuring a stable detection baseline.

[0029] In one embodiment, the step of performing hierarchical screening based on a constructed information mapping table, calculating a priority index by combining hardware qualifications, calibration accuracy, and dynamic parameters, and selecting a unique initial candidate detection probe from all probes includes: S110: Retrieve the multi-dimensional probe basic information mapping table and complete the field classification and isolation. Perform the first round of screening based on hardware batches and repair records, and collect compliant probes to form a first-level qualified candidate set. S111: Read the calibration error and static baseline data from the first-level qualified candidate set, calculate the baseline correction coefficient, eliminate probes with abnormal coefficients, and retain compliant probes to form the second-level candidate subset; S112: The calibration error and baseline correction coefficient of the secondary candidate subset are used for calculation and processing to synthesize the comprehensive accuracy index and sort and select the best to form a third candidate set within a limited range; S113: Rank the dynamic parameters of the probes in the third candidate set, calculate the comprehensive priority index according to the preset weight, and select the unique initial candidate detection probe by comparison according to the index rules.

[0030] In this embodiment, the first step is to retrieve the multi-dimensional probe basic information mapping table and complete field classification and isolation. A first round of screening is then conducted based on hardware batches and repair records, resulting in a primary qualified candidate set of compliant probes. Using the previously deployed P01-P08 conventional probes and P09 inspection probe, the established multi-dimensional information mapping table is retrieved, and the hardware batches, repair records, calibration parameters, baseline data, and five dynamic parameters are classified and stored separately. Hardware batches are converted into four-digit year-month codes, and the screening threshold is set to the batch code corresponding to six months prior to the current time. Repair records are identified using flag bits: flag bit 0 indicates no repair, and flag bit 1 indicates repair. Among the field probes, P03, P04, and P05 have batch codes greater than the threshold and a repair flag bit of 0. The remaining probes have outdated batches or repair markings and are all directly eliminated. P03, P04, and P05 are collectively grouped to form a primary qualified candidate set, completing the first round of hardware qualification screening and defining the scope of compliant probes for subsequent accuracy screening.

[0031] Next, calibration error and static baseline data are read from the primary qualified candidate set to calculate the baseline correction coefficient. Probes with abnormal coefficients are eliminated, and compliant probes are retained to form the secondary candidate subset. The factory calibration error and static baseline deviation values ​​of the three probes in the primary set are read, and the baseline correction coefficient is calculated according to a fixed formula, with the compliance range limited to 0.9 to 1.1. Among them, the baseline correction coefficient of P03 is 1.02, and that of P04 is 1.05, both within the compliance range; the correction coefficient of P05 is 1.13, which exceeds the set range and is directly eliminated. Probes P03 and P04 are retained to form the secondary candidate subset, completing the second layer of screening of accuracy parameters and eliminating unqualified probes with excessive baseline offset. Next, the calibration error and baseline correction coefficient of the secondary candidate subset are used for calculation to synthesize a comprehensive accuracy index, which is then sorted and selected to form a third candidate set within a limited range. The factory calibration error is multiplied by the baseline correction coefficient to obtain the corrected calibration error, which is then added to the static baseline deviation value to generate the comprehensive accuracy index. The overall accuracy index of P03 is 0.38, and that of P04 is 0.65. Arranged in ascending order of accuracy, probes with lower overall accuracy indices are retained, leaving only P03 in the third candidate set. The third candidate set contains only a single probe, eliminating the need for further elimination. This completes the final selection based on accuracy, identifying the single candidate with the highest accuracy.

[0032] Next, the dynamic parameters of the probes in the third candidate set are ranked, and a comprehensive priority index is calculated according to preset weights. A unique initial candidate probe is selected based on the index rules. Five dynamic parameters are selected: historical service time, cumulative number of tests, environmental exposure risk score, fault interval duration, and power supply voltage margin, with preset weights of 25%, 20%, 25%, 15%, and 15%, respectively. Each parameter is ranked according to fixed rules, and a comprehensive priority index is calculated by weighted summation. P03 ranks highly in all dynamic parameters, resulting in a comprehensive priority index of 18.5, the lowest among all currently screenable probes. P03 is directly selected as the unique initial candidate probe according to the rules; if the indices are the same, the shortest historical service time is used as the fallback criterion. This step, through multi-level fixed rule screening and weighted index calculation, objectively completes the probe selection, providing a definite hardware object for subsequent probe status verification and commissioning, ensuring that the initial configuration of the entire dew point inspection process is standardized, quantifiable, and closed-loop.

[0033] In one embodiment, the step of activating the dedicated inspection probe and simultaneously collecting dew point detection data corresponding to the conventional detection probe and the inspection probe when the preset inspection cycle is reached includes: S20: Runs a timed task in the background and matches the built-in inspection cycle parameters. The cycle is determined by the system timestamp. Once the cycle conditions are met, an inspection trigger event is generated, the dedicated inspection probe is turned on, and the preheating process begins. S21: After the preheating is completed, the sampling sequence numbers of the two sets of probes are reset and cleared to zero. The inspection probes are unlocked, initialized, handshaked, and self-checked to maintain the uninterrupted online acquisition link of the regular detection probes. S22: Send a synchronous trigger enable signal, define the acquisition time window and allocate an independent communication time slot, issue a unified sampling scheduling command, control the two probes to carry out sampling in parallel under the same sequence and working conditions and mark the sampling sequence number; S23: Based on the sampling sequence number, complete the point-to-point pairing and encapsulation of the two sets of data, collect them into the special data buffer for point inspection, complete the timing alignment according to the queue rules, store the normalized data into the corresponding comparison partitions, and restore the regular probe data queue state.

[0034] In this embodiment, the system first runs a timing task in the background and matches the built-in inspection cycle parameters. The cycle is determined using the system timestamp. Once the cycle condition is met, an inspection trigger event is generated, the dedicated inspection probe is activated, and the system enters a warm-up process. This implementation uses the previously determined conventional probe P03 and dedicated inspection probe P09. The system's built-in inspection cycle default parameter is fixed at 25,920,000 seconds. The timing task runs independently in the background and does not occupy main business process resources. The system reads the built-in system clock timestamp in real time, records the timestamp of the previous round of inspection completion, and calculates the difference between the two timestamps in real time. When the time difference is greater than or equal to 25,920,000 seconds, the system sets the inspection trigger flag, generates an inspection trigger event, and writes it to the task queue. The system sends an activation command code 0x02 to the inspection probe P09. After receiving the probe's feedback status code 0xBB, it enters a fixed 60-second warm-up process. During the warm-up period, the P03 conventional detection link remains operational, ensuring uninterrupted real-time dew point acquisition and preparing the hardware and operating conditions for subsequent dual-probe synchronous sampling.

[0035] Secondly, after the warm-up period, the sampling sequence numbers of both sets of probes are reset and uniformly cleared. Unlocking initialization, link handshake, and status self-checks are performed on the inspection probes, maintaining the uninterrupted online acquisition link of the regular inspection probes. After the 60-second warm-up process, the system sends a sampling sequence number reset command 0x04 to the regular probe P03, and simultaneously clears the internal counter register of the inspection probe P09, setting the sampling start sequence numbers of both sets of probes to 0 and archiving the configuration snapshot. The system sends software unlocking and initialization commands to P09, defining an independent isolated communication zone and shielding it from signal interference from other devices on the bus. Identity verification is completed through an eight-bit random number challenge and key-encrypted response, establishing a dedicated point-to-point communication link. The warm-up duration and acquisition time window are configurable adaptive parameters. During normal inspection, the original online acquisition link of the regular inspection probes remains uninterrupted, only shutting down the original probe acquisition channel during abnormal switching phases. A dual-redundant frame structure is used to complete link handshake, communication rate and verification rule matching, initiate high-priority arbitration on the bus, and obtain exclusive bus access. After completing the humidity sensor sensitivity calibration, temperature compensation parameter detection, and output linearity verification in sequence, P09 enters the ready-to-collect state after all self-test items pass, maintaining the original online acquisition link of P03 without interruption throughout the process.

[0036] Next, a synchronization trigger enable signal is sent, a data acquisition time window is defined and an independent communication time slot is allocated, and a unified sampling scheduling command is issued to control the two probes to conduct parallel sampling under the same sequence and operating conditions, and to mark the sampling sequence number. The system sends a 10-millisecond high-level hardware trigger pulse to the synchronization trigger module, which automatically goes low to take effect. A fixed data acquisition time window is defined, and independent dedicated communication time slots are allocated to P03 and P09, and a unified sampling scheduling command is issued. The sampling frequency of both probes is set to once per second, with 100 sampling points continuously acquired in a single acquisition, using a unified clock edge triggering method. P03 and P09 are controlled to start sampling in parallel at the same time and under the same field conditions. Each set of sampled data is accompanied by a sequentially increasing sampling sequence number to ensure that the timing scales of the two data streams are completely aligned, eliminating acquisition errors caused by timing deviations and differences in operating conditions. Then, the two sets of data are paired and encapsulated point-to-point according to the sampling sequence number, collected into a dedicated data buffer for inspection, and time-aligned according to the queue rules. The normalized data is then stored in the corresponding comparison partitions and the regular probe data queue state is restored. The system pairs data points with the same sequence number (P03 and P09) one by one according to the sampling sequence number from smallest to largest, and encapsulates the data by adding a unified window identifier and time sequence number. The encapsulated data is then uniformly collected into a dedicated data buffer for inspection, and time sequence alignment and calibration are performed according to the first-in, first-out (FIFO) queue rule. The aligned P03 data is stored in the regular comparison partition, and the P09 data is stored in the inspection comparison partition, with each partition storing a fixed 100 data points. After data regularization and storage are completed, the P03 regular inspection data queue is unfrozen, normal data writing logic is restored, and the buffer lock is released. If the number of sampling points from the two probes does not match or time sequence alignment fails, the current round of inspection is invalidated, awaiting the next cycle trigger. This step achieves time sequence alignment and regularization of the two probe data through precise cycle triggering, probe preheating self-test, and simultaneous sampling under the same sequence and operating conditions, providing a reliable standard data source for subsequent difference comparison.

[0037] In one embodiment, after the preheating is completed, the sampling sequence numbers of the two sets of probes are reset and uniformly cleared. Unlocking initialization, link handshake, and status self-check are performed on the inspection probes to maintain the uninterrupted online acquisition link of the regular detection probes. This includes the following steps: S210: After the preheating is completed, a sampling sequence number reset command is issued. After receiving the probe confirmation feedback, the two sets of probe counting units are cleared, the unified sampling start sequence number is fixed and the configuration snapshot is archived, and the initial counting benchmark is locked. S211: Sends an unlocking and initialization command to the dedicated inspection probe, divides the independent isolated communication zone and shields bus interference, completes identity verification through identity challenge and response interaction, and establishes a point-to-point dedicated communication link; S212: Follow the fixed frame format and interaction timing to carry out link handshake, complete the matching and verification of communication-related parameters, mark the valid link and initiate a bus arbitration request to obtain bus-specific access rights; S213: By issuing status self-test commands through the established communication link, the probe's core parameters and output performance are verified item by item, parameter normalization and communication protocol adaptation configuration are performed, and the working parameters are solidified so that the inspection probe enters the ready-to-collect state.

[0038] In this embodiment, firstly, after the preheating period, a sampling sequence number reset command is issued. After receiving confirmation feedback from the probes, the counting units of both probes are reset to zero, a unified sampling start sequence number is fixed, and a configuration snapshot is archived to lock the initial counting benchmark. Using the conventional detection probe P03 and the dedicated inspection probe P09 described earlier, after a 60-second preheating period, the system issues a sampling sequence number reset command code 0x04. P03 and P09 return confirmation frames 0xCC respectively. Upon receiving the feedback, the system synchronously resets the internal sampling sequence number counting units of both probes to zero. The sampling start sequence numbers of both probes are uniformly set to 0, and a configuration snapshot is generated and archived to lock the initial counting benchmark, preventing any changes. Throughout the process, the original online acquisition link of P03 remains uninterrupted; only the sampling count sequence number is uniformly reset to zero, establishing a counting benchmark for subsequent pairing of data points with the same sequence number by dual probes. Secondly, an unlock initialization command is sent to the dedicated inspection probe to divide an independent isolated communication zone and shield bus interference. Identity verification is completed through identity challenge and response interaction, establishing a point-to-point dedicated communication link. The system issues an unlock initialization command code 0x02 to P09, waits for valve feedback frame 0xBB to confirm physical opening. The system allocates an independent isolated communication zone for P09, sets a dedicated access whitelist, and shields the signal interference from other probes and devices on the bus. The system generates an eight-bit random number as challenge data and issues an identity challenge command code 0x20; P09 calls a preset key for encryption and returns a response frame 0x21. After successful decryption and verification, the system establishes a dedicated point-to-point communication link between P09 and the main control unit, isolating external communication interference and ensuring unique and reliable command and data transmission.

[0039] Next, following a fixed frame format and interaction sequence, a link handshake is performed to complete the matching and verification of communication-related parameters. A valid link is marked, and a bus arbitration request is initiated to obtain exclusive bus access rights. The system, according to a preset fixed frame format and a master-slave frame transmission sequence with a five-millisecond delay, performs byte-by-byte data comparison and interaction with P09. The device hardware address, communication baud rate, and data verification rules are matched sequentially. After all matches are correct, the current communication link is marked as a valid link, and the handshake record is saved. The system initiates a high-priority arbitration request to the bus, with a priority code of 0x50. Upon receiving the request, P09 sends back an arbitration success frame (0x53), exclusively securing bus access rights to prevent other devices from preempting the channel and reserving dedicated bus resources for subsequent self-test command issuance and data acquisition. Then, a status self-test command is issued through the established communication link to verify the probe's core parameters and output performance item by item, perform parameter normalization and communication protocol adaptation configuration, and solidify the working parameters, putting the inspection probe into a ready-to-collect state. The system issues a self-test command code 0x54 to verify the humidity sensor's sensitivity calibration coefficient, temperature compensation algorithm parameters, and output linearity segmentation index in sequence. After all items pass verification, P09 sends a self-test pass frame 0x55. The system then issues a preprocessing configuration command code 0x86, setting fixed rules for three retransmissions of error reports, data output with checksum, and a two-second communication timeout. After P09 writes the configuration, it sends a completion frame 0x87, solidifying the working parameters to local storage. The inspection probe officially enters the ready-to-collect state, completing all preparatory work for simultaneous synchronous sampling by dual probes. This step ensures reliable probe readiness and does not affect continuous testing by regular probes by using a unified sampling counting benchmark, isolating the communication link, and performing identity handshake self-tests to shield against bus interference.

[0040] In one embodiment, the step of performing a difference calculation on two sets of dew point detection data, comparing the calculation result with a preset judgment threshold, and obtaining a probe working status judgment result includes: S30: Read two sets of dew point detection data with time alignment, complete precise pairing based on synchronization identifier, remove abnormal and invalid data points and define judgment data window, construct standard comparison dataset, and generate signed difference sequence by point-by-point operation; S31: Divide the dataset into segments and sliding windows, calculate the deviation statistics segment by segment, count the proportion of outliers and the distribution of the sign of the difference in the window, recursively calculate the cumulative sum sequence and complete the statistics of out-of-bounds points. S32: Retrieve the system's preset constraint thresholds and judgment limits, and conduct comprehensive cross-comparison and verification of segmented deviation, overall deviation, data fluctuation, sign consistency, and cumulative and out-of-bounds situations; S33: Based on the combination and matching relationship of multi-dimensional verification indicators, distinguish between the normal state of the probe and different abnormal types, and generate a fixed probe working status judgment result.

[0041] In this embodiment, the two sets of dew point detection data are 100 time-aligned dew point data points synchronously collected by the conventional detection probe P03 and the dedicated inspection probe P09 mentioned earlier. These data are completely consistent with the data used for synchronous sampling mentioned earlier, with each set containing 100 dew point temperature values. Taking a typical industrial site as an example, the output sequence of P03 is [-10.2℃, -10.5℃... a total of 100 points], and the output reference sequence of P09 is [-10.1℃, -10.3℃... a total of 100 points]. Based on the sampling sequence number and synchronization identifier, each point is paired one-to-one. Three invalid isolated data points with frame errors or value jumps are directly removed, and the remaining 97 valid points are designated as a fixed judgment data window, forming a standard comparison dataset. Point by point, the P09 value is subtracted from the P03 value to generate a 97-point signed difference sequence, such as a difference of -0.1℃ for the first point and -0.2℃ for the second point. This difference sequence is the core foundational data for subsequent deviation statistics and threshold comparison, ensuring the consistency of the closed loop of the data link. Next, the preset fixed parameters are used: a sliding window length of 20 points and a step size of 10 points, dividing the entire sequence into 4 independent segments of 25 points each. The 97-point difference sequence is then divided into local windows and fixed segments, and the median and average deviation values ​​are calculated for each segment. The sliding window length, step size, and number of points in each segment are preset configurable parameters. Taking the first 20-point sliding window as an example, the number of outliers exceeding the preset ±2℃ threshold is counted as 2, and the local anomaly ratio is calculated. The number of positive and negative points in the difference sequence is counted to obtain the sign consistency ratio. The system synchronously generates a first-level signed cumulative sum and a second-level absolute value cumulative sum sequence. It counts the number of out-of-bounds points exceeding the first-level threshold of ±10℃ and the second-level threshold of 0~15℃ according to a fixed traversal rule. It retains all the original values ​​of segmented statistics, window statistics, cumulative sums and out-of-bounds points, providing quantitative data support for multi-dimensional comparison.

[0042] The system then directly calls the fixed configuration parameters: dew point judgment threshold ±2℃, first-level cumulative sum threshold ±10℃, second-level cumulative sum threshold 0~15℃, segmented cumulative sum threshold ±8℃, and three thresholds for the number of out-of-limit windows. It then compares the median deviation of each segment and the overall average deviation with the thresholds item by item, while simultaneously verifying the data fluctuation dispersion, sign consistency ratio, cumulative sum, and number of out-of-limit points. Each item is checked against the threshold without any additional parameter adjustments; only fixed rule matching is performed to lock all out-of-limit indicators and the number of out-of-limit windows. Then, following the fixed judgment logic described above, taking P03 detection data as an example, when the segmented deviation, overall deviation, and data fluctuation are all within the threshold, the number of out-of-range windows is less than 3, and the sign consistency and out-of-bounds points do not exceed the threshold, the probe is judged to be working normally, and the status code 0x00 is output; when the number of out-of-range windows is ≥3, and the sign ratio and the number of out-of-bounds points exceed the standard simultaneously, it is judged to be a systematic cumulative drift anomaly, and the status code 0x09 is output; the remaining matching relationships correspond to random local anomalies and complex anomalies, respectively, and the corresponding fixed status codes are output. The first-level signed cumulative sum is set as the main judgment criterion, and the second-level absolute value cumulative sum and segmented deviation are used as auxiliary judgment criteria. When there is a conflict between local and global judgments, the global signed cumulative sum shall prevail. This step accurately identifies the probe's working condition and anomaly type through data cleaning and pairing, segmented sliding statistics and cumulative sum analysis, and multi-dimensional cross-threshold comparison.

[0043] In one embodiment, the steps of segmenting the dataset and dividing it into sliding windows, calculating the deviation statistics segment by segment, statistically analyzing the proportion of outliers and the distribution of the difference sign within the window, recursively calculating the cumulative sum sequence, and completing the statistics of out-of-bounds points include: S310: Read the signed difference sequence and initialize the cumulative sum benchmark, allocate a dedicated storage area for the sequence, perform validity checks on the difference data, remove invalid data, and complete the classification labeling according to positive and negative types; S311: Perform cumulative operations according to positive and negative categories, independently record the cumulative results of the two types of differences, and synchronously recursively generate the first-level cumulative sum corresponding to the original difference and the second-level cumulative sum sequence corresponding to the absolute value of the difference; S312: Divide the difference data into segments as a whole, establish an independent cumulative sum sequence for each segment, calculate the deviation statistics for each segment, count the outlier points in each segment and calculate the proportion of local outliers, and form a segmented cumulative sequence. S313: Iterate through the cumulative sums and segmented subsequences at each level, compare and screen outbound points and abnormal segments, summarize various statistical indicators, and store the statistical results and each cumulative sum sequence into their respective independent storage areas.

[0044] In this embodiment, the first signed difference sequence is the same as the 97 dew point difference sequence generated by comparing the conventional probe P03 with the inspection probe P09 mentioned earlier, which is completely consistent with the output data of the previous step. The system initializes the cumulative sum baseline value to 0 and allocates two dedicated storage areas to store the first-level cumulative sum and the second-level cumulative sum sequences, with a storage capacity matching 100 data points. Taking the field measurement data as an example, the difference sequence includes values ​​such as -0.1℃, 0.3℃, 2.2℃, and -0.2℃. Data validity verification is performed, and invalid outlier points exceeding the reasonable range are directly removed, retaining 95 valid difference data. Valid data are classified and labeled according to positive and negative attributes, with positive values ​​marked as positive differences and negative values ​​marked as negative differences. This step unifies the data baseline and storage space, filters out erroneous data, and defines a compliant and valid data source for subsequent cumulative calculations. Secondly, the 95 retained valid difference data are used as the calculation objects. Without canceling out positive and negative values, the positive and negative difference accumulation results are calculated separately. The first-level cumulative sum is calculated by recursively accumulating the original signed differences point by point, while the second-level cumulative sum is calculated by recursively accumulating the absolute values ​​of the differences point by point. For example, if the first point difference is -0.1℃, the first-level cumulative sum is recorded as -0.1℃, and the second-level cumulative sum is recorded as 0.1℃; if the second point difference is 0.3℃, the first-level cumulative sum is recorded as 0.2℃, and the second-level cumulative sum is recorded as 0.4℃. This process is repeated point by point to generate a complete 95-point first-level cumulative sum sequence and a 95-point second-level cumulative sum sequence. The two types of sequences are stored independently without interference, providing two sets of standard statistical sequences for subsequent threshold exceeding judgment.

[0045] Next, using fixed parameters, the 95-point difference sequence is divided into four fixed segments of 25 data points each. A separate sub-storage area is allocated for each segment, creating an independent segmented cumulative sum sequence. The median and average deviation of the dew point deviation are calculated segment by segment. Taking the first 25-point segment as an example, the number of abnormal points exceeding the preset threshold of ±2℃ is counted as 3, and the proportion of local anomalies is calculated as 12%. Segmented cumulative final values ​​and anomaly statistics are generated for each segment, summarizing to form a complete segmented cumulative sequence, thus solidifying the deviation characteristics and anomaly distribution information of each segment's data. Then, the system retrieves fixed thresholds: Level 1 cumulative sum ±10℃, Level 2 cumulative sum 0~15℃, and segmented cumulative sum ±8℃. The Level 1 and Level 2 cumulative sum sequences are traversed point by point, counting the number of points exceeding the threshold; the four segmented cumulative sum sub-sequences are traversed segment by segment, comparing whether the segmented final values ​​exceed the limits, and marking the number of abnormal segments. All indicators, including the number of Level 1 out-of-bounds points, the number of Level 2 points exceeding the upper limit, the number of segmented abnormal segments, and the proportion of local anomalies, are summarized. All statistical indicators, first-level cumulative sum sequences, second-level cumulative sum sequences, and four-segment segmented cumulative sum sequences are written to the system's preset independent storage partitions. This step quantifies the deviation characteristics through difference data classification verification, two-level cumulative sum recursion, and segmented statistics, providing a quantitative basis for accurate probe fault determination.

[0046] In one embodiment, the step of performing cumulative operations according to positive and negative categories, independently recording the cumulative results of the two types of differences, and synchronously recursively generating a sequence of first-level cumulative sums corresponding to the original differences and second-level cumulative sums corresponding to the absolute values ​​of the differences includes: S3110: Read the signed difference sequence and set the initial value of the cumulative sum, create a two-level cumulative sum sequence buffer and write the initial value, perform validity verification on the difference data, and divide the positive and negative difference classification bits; S3111: Separate and classify the valid difference data according to positive and negative attributes and record the cumulative results independently. Simultaneously establish the original difference buffer and the absolute difference buffer, store the corresponding difference data into them respectively, and generate two sets of buffers simultaneously. S3112: Read the first point data of the two sets of buffers, calculate the first-level cumulative sum and the second-level cumulative sum of the first point respectively in combination with the initial value, and write the calculation results into the corresponding storage location of the two-level cumulative sum buffer; S3113: Iterate through each level of cumulative sum according to the unified recursive rule, and use a dual-channel parallel computing method to synchronously complete data transmission and iterative processing until all point operations are completed, and synchronously generate a complete two-level cumulative sum sequence.

[0047] In this embodiment, the signed difference sequence here first uses the 95-point dew point signed difference sequence generated by the conventional detection probe P03 and the spot inspection probe P09 mentioned earlier, which is completely consistent with the data used in the previous step of statistics. The system sets the initial value of the cumulative sum to 0, creates a first-level cumulative sum buffer and a second-level cumulative sum buffer, and configures the buffer capacity for 100 data points in each buffer. The initial value of 0 is written to the first address of each of the two buffers. The validity of each of the 95-point difference data is verified, and invalid outlier points that exceed the dew point range are removed, while all compliant and valid data are retained. Positive and negative classification markers are set for each valid data point. A difference greater than 0 is marked as positive, a difference less than 0 is marked as negative, and a difference of 0 is marked separately. This step fixes the initial calculation benchmark, allocates a dedicated storage area, completes data compliance screening and positive and negative classification marking, and defines the standard input data source for the two-level cumulative sum recursive calculation. Next, the 95 valid difference data that have passed the verification are split and classified according to the set positive and negative markers. Positive differences and negative differences are stored separately and do not overlap or cancel each other out. Simultaneously, a raw difference buffer and an absolute difference buffer are created, with buffer capacities matching 95 valid data points. The signed raw differences are sequentially written to the raw difference buffer, and the absolute values ​​of each difference are sequentially written to the absolute difference buffer. Taking the first three points of the sequence as an example, the raw differences are -0.1℃, 0.3℃, and -0.2℃, and the absolute differences are 0.1℃, 0.3℃, and 0.2℃, respectively. The two buffers are filled and stored synchronously according to the data point order, completely preserving the original data of the raw and absolute differences, providing two fixed input data sources for dual-channel parallel computation.

[0048] Next, the system sequentially reads the first point value of the original difference buffer (-0.1℃) and the first point value of the absolute difference buffer (0.1℃). Using the global initial value of 0 as the calculation base, the first-level cumulative sum equals the initial value plus the original difference, resulting in -0.1℃; the first-level cumulative sum equals the initial value plus the absolute difference, resulting in 0.1℃. -0.1℃ is written to the second address of the first-level cumulative sum buffer, and 0.1℃ is written to the second address of the second-level cumulative sum buffer. Strictly adhering to address offset rules, the storage location of the first-point calculation result is fixed, establishing the starting calculation node for the two-level cumulative sum recursive calculation. Then, a unified recursive rule is set: the Nth point's first-level cumulative sum equals the value of the previous-level cumulative sum buffer plus the Nth point value of the original difference buffer; the Nth point's second-level cumulative sum equals the value of the previous-second-level cumulative sum buffer plus the Nth point value of the absolute difference buffer. A dual-channel parallel calculation is configured: channel one handles the first-level cumulative sum iteration calculation and data transmission, and channel two handles the second-level cumulative sum iteration calculation and data transmission. Both channels run independently. The calculation is iteratively performed on each of the 95 valid data points until the last point is calculated. After the calculation, the first-level cumulative sum buffer generates a complete original difference cumulative sequence of 95 points, and the second-level cumulative sum buffer generates a complete absolute value cumulative sequence of 95 points. The two sets of sequences are synchronously and permanently stored, directly used for subsequent threshold out-of-bounds traversal and segmented statistics. This step efficiently generates two-level cumulative sum sequences through difference classification and labeling, dual-buffer storage, and dual-channel parallel recursion, ensuring accurate data calculation and time synchronization.

[0049] In one embodiment, the step of activating a backup probe to replace the regular dew point detection and restoring the original abnormal detection probe and the inspection probe to a sealed and closed state if the comparison result exceeds the threshold includes: S40: After determining that the threshold comparison result is abnormal, lock the two types of probe device identifiers and establish a dedicated control link to isolate external signal interference. After reading the status flag bit, issue shutdown commands in sequence and receive corresponding feedback frames. S41: Verify the timing information of the feedback frame to confirm that the instruction is executed effectively, issue a collection termination instruction and shut down the collection channel and communication port, clear the collection buffer, and reissue the compliant shutdown instruction after verifying the communication link through test frame interaction. S42: Issue sampling pause, moderate shutdown and complete shutdown control commands in stages, update the sealed status flag after receiving execution feedback, initiate bus release request and set the communication port to isolated state; S43: Solidify the sealing status configuration of the two types of probes and complete the status verification, disable the relevant control and acquisition interfaces, reset the buffer, address pointer and counting resources, send the start command to the backup probe, and complete the switching over of the routine dew point detection task.

[0050] In this embodiment, the anomaly detection target is the conventional detection probe P03 mentioned earlier, and the dedicated inspection probe is P09, which are completely consistent with the probes used in the data comparison and judgment mentioned earlier. After multi-dimensional threshold comparison, the dew point difference of P03 exceeds the preset threshold of ±2℃, and it is judged as a detection anomaly. The system immediately locks the hardware number of the devices P03 and P09, establishes a dedicated point-to-point control link between the two, and shields the signal interference from other devices on the industrial bus. The status flag bits of the two probes are read, with flag bit 1 representing open and flag bit 0 representing sealed closed. At this time, the status flag bits of P03 and P09 are both 1. The system sends the shutdown command 0x03 and 0x0E in a fixed order, while waiting for the probe to return the corresponding feedback frames 0xCC and 0xEE, locking the target device and communication link, and establishing a dedicated hardware connection for subsequent valve closing control. Next, the system records the timestamps of the two feedback frames received, and verifies that the difference between the timestamps is less than 100ms, and judges that the two-level shutdown command is valid. The acquisition termination command is issued to P03 and P09 to directly shut down the acquisition channels of the two probes and the bus communication port. The system clears the real-time acquisition buffer and historical sampling data buffer inside both probes. It intentionally sends test frames with incorrect checksums; P03 and P09 detect the errors and return rejection frames 0x82, completing the communication link activity verification. After successful link verification, a standard compliant shutdown command is reissued to ensure the communication link is normal and reliable, preventing probe status loss or communication failure.

[0051] The system then issues control commands according to a three-stage fixed process. The first stage issues a sampling pause command (0x10), maintaining the pause for 2 seconds to stop data acquisition. The second stage issues a valve partial closure command (0x11), reducing the valve opening to 50% and maintaining this for 3 seconds. The third stage issues a complete closure command (0x03), setting the valve opening to 0. Both probes execute these commands sequentially and return confirmation feedback frames. The system updates the P03 and P09 status flags to 0, marking them as sealed and closed. A release request command (0x18) is issued to the bus. Upon receiving the bus release confirmation frame (0x19), the communication ports of both probes are set to a high-impedance isolated state, removing them from bus scheduling. Finally, the system issues a configuration solidification command (0x16), writing the sealed and closed states of P03 and P09 to non-volatile memory and receiving a solidification confirmation frame (0x17), completing the status solidification verification. The control and acquisition interfaces of both probes are permanently disabled, the data buffer is cleared, the address pointer is reset, and the sampling sequence counter is cleared. The system sends an activation command 0x01 to the first standby probe P04 in the circular queue. After P04 returns a status code 0xAA, the status flag is updated to 1, officially taking over the routine dew point detection task on site. This step reliably isolates abnormal probes by locking the probe link, performing graded command verification and phased shutdown, and achieves seamless switching of standby probes to ensure uninterrupted detection.

[0052] In one embodiment, the steps of locking the identifiers of the two types of probe devices and establishing a dedicated control link after the threshold comparison result is abnormal, isolating external signal interference, reading the status flag bit, issuing shutdown commands sequentially level by level, and receiving corresponding feedback frames include: S401: Lock onto the target probe device identifier, establish a dedicated control link and isolate external signal interference, read the status flag bit and complete the validity verification, and create a command interaction waiting record table; S402: Sends the first round of shutdown commands to the two types of probes and configures the exclusive serial number, opens the dedicated receiving channel, starts the listening mechanism and activates the receiving interrupt to capture the feedback frame; S403: After capturing the feedback frame, turn off listening and interrupt, record time information, verify the relevant identifiers and codes of the feedback frame, update the record table and send a completion signal after matching; S404: After receiving the completion signal, it sends the next round of shutdown command, reconfigures the interrupt conditions and restarts the listening, completes the feedback frame verification, marks the execution status and confirms that the two-level command interaction is complete.

[0053] In this embodiment, the target probes are the conventional detection probe P03 and the dedicated inspection probe P09, which were previously identified as abnormal. These probes are identical to those used in the previous anomaly detection step. The system locks the fixed hardware device numbers of P03 and P09, establishing a dedicated control communication link between them to shield against signal interference from other probes and peripherals on the industrial fieldbus. The system reads the status flags of the two probes, setting flag 1 to probe open and flag 0 to probe sealed closed. Currently, both the status flags of P03 and P09 are 1, verifying that the status is valid and a shutdown operation can be performed. The system creates a new instruction interaction waiting record table, which includes five fixed fields: instruction type, expected feedback frame, instruction sending timestamp, feedback receiving timestamp, and execution completion status. This provides a standardized record for subsequent two-level shutdown instruction interactions. Next, the system simultaneously sends the first-round shutdown instruction 0x03 to P03 and P09, configuring a unique serial number 0025 for this instruction. The dedicated receiving channel numbered 0x10 is activated, receiving only feedback data from the two target probes. The bus data monitoring mechanism is initiated, and a high-priority receive interrupt is activated. The interrupt trigger condition is set to match frame header 0×CC. The system is set to a timeout of 500 milliseconds, waiting for probe response feedback. During this period, it does not respond to bus messages from other unrelated devices, but only captures and listens for feedback frames from P03 and P09 to ensure accurate one-to-one interaction of the first round of shutdown commands. A command response timeout threshold is set. If no feedback frame is received within the timeout period, the system will automatically retransmit up to 2 times. If there is still no response, the probe communication is directly determined to be faulty and marked as invalid.

[0054] After the receive interrupt captures a feedback frame with a header of 0xCC, the system immediately disables the bus listening mechanism and disables the receive interrupt enable. It records the instruction transmission timestamp and the feedback reception timestamp, calculating the interval between them. The system sequentially verifies the device address, frame type encoding, and preset sequence number of the feedback frame, while simultaneously calculating the checksum according to rules and comparing it bit-by-bit with the checksum field built into the frame. Once all identifiers, encodings, and checksums match, the system marks the instruction execution completion status as confirmed in the instruction interaction wait record table and sends a first-round instruction interaction completion signal to the system main program, solidifying the first-round instruction execution result. Upon receiving the first-round completion signal, the system issues a second-round shutdown instruction 0x0E to P03 and P09, reconfiguring a new dedicated sequence number 0026. The receive interrupt trigger condition is reconfigured to match the frame header 0xEE, and the bus listening and 500ms timeout timer are restarted. After capturing the target feedback frame, the system again performs timestamp recording, sequence number comparison, frame encoding verification, and checksum verification. After verification, the system marks the completion status of the second round of command execution in the command interaction waiting record table. The system determines that the first and second rounds of two-level shutdown command interactions have been completed in compliance with regulations, and locks the command response status of the two probes. This step avoids bus interference and ensures accurate and reliable shutdown command interaction by locking the dedicated link of the probe, issuing commands in a hierarchical manner with serial numbers, and monitoring and verifying interrupts.

[0055] In one embodiment, the steps of sending the first round of shutdown commands to the two types of probes and configuring dedicated serial numbers, opening a dedicated receiving channel, starting the monitoring mechanism, and activating the receiving interrupt to capture feedback frames include: S4020: Sends a first-round shutdown command with a sequence identifier to both types of probes, opens the dedicated receiving channel, saves the task context, and defines the blocking wait time boundary; S4021: Blocks the original execution flow of the program, calls the task suspension and semaphore wait functions, and moves the current task into the blocking queue and the semaphore wait queue; S4022: Masks system-irrelevant interrupts and low-priority task scheduling, enables dedicated receive interrupts, releases semaphores and wakes up blocked processes after matching the target feedback frame; S4023: After the process resumes running, verify the semaphore status and timing information, clear the task suspension flag, close the dedicated receive channel, destroy the semaphore and release the blocking wait state.

[0056] In this embodiment, the two types of probes used here are the conventional detection probe P03 and the dedicated inspection probe P09, which were previously identified as abnormal. These are identical to the probes used in the previous instruction interaction. The system sends a first-round shutdown instruction 0x03 to P03 and P09, carrying a fixed sequence identifier 0025 to ensure traceability. The dedicated receiving channel with channel number 0x10 is opened, receiving only feedback frames from the target probe. The current program counter, registers, variable addresses, and other task contexts are saved, and a fixed blocking wait time boundary of 500 milliseconds is defined; if this timeout occurs, the instruction interaction is considered a failure. This step fixes the instruction and channel parameters, saves the runtime environment, and sets the timeout boundary, providing the basic conditions for blocking wait and feedback capture. Next, the system directly blocks the original main program execution flow, preventing the execution of other acquisition, calculation, and communication tasks. The operating system task suspension function is called, and the binary semaphore is initialized with an initial value of 0. The semaphore wait function is called, and the current task requests to acquire the semaphore. Because there are no semaphore resources, the system moves this instruction interaction task into the kernel blocking queue and simultaneously into the semaphore wait queue. The task priority is set to the highest level to prevent it from being preempted by other tasks. This step pauses the program flow according to fixed rules, completes the task state switch, prepares for waiting for probe feedback frames, establishes a dedicated receiving and listening channel, blocks bus-irrelevant messages and interference from low-priority tasks, and only parses valid response frames from the target probe.

[0057] Next, the system blocks irrelevant interrupts from the serial port, timers, and other probes, prohibits low-priority task scheduling, and enables only the dedicated receive interrupt. The interrupt trigger condition is frame header 0xCC and device address matching P03 and P09. The probe returns feedback frame 0xCC, and the system matches the frame header, address, and sequence identifier. Immediately, a semaphore release operation is performed, setting the semaphore value to 1. The kernel automatically wakes up the current instruction interaction task from the blocking queue and restores the task's execution privileges. This step blocks interference sources, responds only to target feedback, and completes the semaphore and process state switch. After the task resumes execution, the semaphore status is read as released, and the feedback reception timestamp is checked; the time interval is 120 milliseconds, not exceeding the 500 millisecond waiting boundary. The timing information is verified as valid, the task suspension flag is cleared, and the task state is set to ready. The 0x10 dedicated receive channel is closed, stopping the reception of target probe data. The semaphore used in this interaction is destroyed, kernel resources are released, and the program is released from its blocking state. The original program execution flow is restored, and the completion of the first round of shutdown instructions is recorded. This step uses task suspension, semaphore blocking and waiting, and dedicated interrupt listening to shield irrelevant interference, accurately capture feedback frames, and improve the stability and real-time performance of instruction interaction.

[0058] In one embodiment, the step of triggering the inspection cycle in successive cycles, comparing data between the inspection probe and the currently used detection probe each time, and outputting a fault indication signal when all reusable detection probes are abnormal, includes: S50: Construct a reusable probe circular queue and initialize it according to priority, independently set up dedicated inspection probes, initialize the inspection state machine and bind the probe identifier, and establish a fixed pairing relationship between the in-use probes and the inspection probes after each round of inspection. S51: After periodic triggering, switch the inspection state machine, synchronously drive the paired probe to complete the unified timing acquisition, cache the acquired data and generate inspection logs, and record inspection-related identifiers and data summary information; S52: Extract the two detection data and perform validity verification and format standardization, construct a standardized dataset, perform difference calculation and threshold comparison based on the inspection probe data, switch the state machine and update the inspection log according to the results; S53: When a probe is found to be abnormal, it is removed from the loop queue, and a subsequent probe is used to take over the detection and the relevant probe status is reset. The inspection process is continuously executed in a loop. When the queue of reusable probes is exhausted, a fault prompt signal is output.

[0059] In this embodiment, the reusable probes used here are P01, P02, P04, P05, P06, P07, and P08 from the previous site description, while the dedicated inspection probe is fixed at P09, consistent with the probe numbers in all the previous inspection processes. The system sequentially enqueues the seven reusable probes into a circular queue based on a comprehensive priority index, from low to high, with the currently used probe P04 initially at the head of the queue. The dedicated inspection probe P09 is filed separately and not included in the circular queue. The inspection state machine is initialized to an idle state by default, and the device identifiers of the head probe P04 and the inspection probe P09 are written into the state machine register. After each inspection cycle of 25,920,000 seconds, the system locks P04 and P09, establishing a one-to-one fixed inspection pairing relationship and generating a pairing index ledger. This step completes the queue setup, state machine initialization, and probe binding, defining a fixed benchmark object for cyclic inspections. Secondly, after the inspection cycle timer expires, the inspection state machine switches from the idle state to the ready state. The system triggers a hardware synchronization pulse, driving the in-use probe P04 and the inspection probe P09 to sample simultaneously and in parallel, with a sampling frequency of 1 second per sampling, continuously collecting 100 data points in a single batch. After collection, the two sets of dew point data are stored in a dedicated double buffer for inspection, without interrupting routine on-site testing. The system creates a new inspection log entry, recording the current timestamp, the in-use probe number P04, the inspection probe number P09, and a summary of the average of the 100 data points. The log is archived sequentially by time. This step completes state machine switching, synchronized sampling, and log recording, preserving the original process information for each round of inspection.

[0060] Next, the system reads 100 sets of dew point data aligned with the time sequence of P04 and P09 from the buffer, removes invalid data such as jumps and null values ​​point by point, and standardizes them into a uniform dataset with consistent accuracy. Using the data from the inspection probe P09 as a benchmark, the dew point difference of P04 is calculated point by point, generating a signed difference sequence. The system's preset ±2℃ judgment threshold, cumulative sum threshold, and window anomaly threshold are invoked to perform segmented deviation, sign consistency, and cumulative sum out-of-bounds cross-comparison checks. If the probe is determined to be normal, the state machine returns to the idle state, and the normal judgment result is added to the original inspection log; if an anomaly is determined, the state machine jumps to the switching state, and the log marks the probe's abnormal state, providing a judgment record for probe rotation. Then, if P04 is determined to be a detection anomaly, the system removes P04 from the head of the circular queue, and the queue length is automatically decremented by one. The next probe in the queue, P02, becomes the new head, and an open command is issued to take over the routine dew point detection on site. The abnormal probe P04 and the inspection probe P09 are uniformly restored to the sealed closed state, and the sampling counter and acquisition channel are reset. The system waits for the next inspection cycle to arrive and repeats the pairing acquisition, difference comparison, status determination, and probe rotation process. When all reusable probes in the circular queue are determined to be abnormal and the queue length is zero, the system outputs a fixed-level fault indication signal. This step is linked with the inspection state machine through the probe circular queue, automatically rotating and comparing for maintenance, promptly replacing abnormal probes, and automatically outputting a fault alarm when probes are exhausted.

[0061] In one embodiment, the steps of constructing a reusable probe circular queue and initializing it according to priority, independently setting up dedicated inspection probes, initializing the inspection state machine and binding probe identifiers, and establishing a fixed pairing relationship between in-use probes and inspection probes after each inspection cycle arrives include: S501: Initialize and sort the reusable probe circular queue according to the preset priority, set the head of the queue to the currently used detection probe, create separate files for the special inspection probes and do not include them in the queue, initialize the inspection state machine and write the probe identifier, generate the paired index key value of the two types of probe identifiers through hash operation and establish a mapping association. S502: Search the system probe status register, filter online target probes that are in a comparable state, send session establishment requests to the in-use probe and the inspection probe respectively, both parties create local session records and return responses, and complete the initial binding of session pairing after verifying that the session numbers are consistent. S503: Lock the identification of two types of probe devices and establish a dedicated communication link, query the probe operating status and issue a status transition command. After both types of probes have completed the status update, a mutually exclusive pairing relationship is formed, restricting other probes from intervening in this inspection process. S504: Generate a comparison ledger containing index information, session information, and status information, write the ledger to the associated storage area, set the ledger completion mark, and switch the working and running state of the inspection state machine.

[0062] In this embodiment, the reusable probes are initially designated as P01, P02, P04, P05, P06, P07, and P08, while the dedicated inspection probe is fixed at P09, consistent with the previous configuration. The seven reusable probes are sequentially queued according to a preset priority index from smallest to largest, with the head probe P04 designated as the currently used conventional inspection probe. A separate file is created for P09, which is not added to the circular queue. The inspection state machine is initialized to an idle state, and the hardware identifiers of the used probe P04 and the inspection probe P09 are written to the state machine context register. A hash index value is calculated for P04 and P09 using an identifier byte accumulation and modulo 256 hash algorithm, establishing a one-to-one fixed mapping association between the two hash values ​​as master and slave keys. This step completes queue sorting, probe partitioning, state machine binding, and index mapping, providing a unique identifier for subsequent session pairing. Next, the system searches the probe status register, selecting online, fault-free P04 and P09 that can participate in the inspection comparison. A 16-byte session number, composed of the current timestamp and an eight-digit random number, is generated and sent to both P04 and P09 simultaneously as a session establishment request message. Probe P04 creates a local session record for the primary role and returns a response frame, while probe P09 creates a local session record for the inspection role and returns a response frame. The system compares the session numbers returned by both parties; if the characters match exactly, the session is deemed legitimate, completing the initial pairing and binding of the two probes and locking in the unique identifier of the session for this inspection.

[0063] Next, the system locks the device identifiers of P04 and P09, establishes an isolated dedicated communication link, and shields the bus from interference from other probe signals. It reads the original operating status of the two probes and sends a status transition command 0x30 to both probes, setting the target status number to 3. P04 and P09 execute the status switch and return confirmation frames, updating the status identifier to 3. The system marks the two probes as entering a mutually exclusive pairing state. Other probes on the bus, after reading the status identifier as 3, automatically refuse to access the current inspection process, retaining only P04 and P09 for this round of data acquisition and comparison. The system then generates a fixed-length comparison log, containing the task number, P04 hash index, P09 hash index, 16-byte session number, status transition flags, lock timestamp, and pending comparison status. The entire log data is written to the circular queue associated storage area and the session management storage area. The system sets the logbook generation completion flag and simultaneously switches the inspection state machine from idle to ready state, waiting for synchronous collection of trigger commands. This step establishes a dedicated communication link through priority queues, hash pairing, and session binding, locking the inspection pairing relationship and ensuring the orderly and isolated operation of the inspection process.

[0064] refer to Figure 2 This application also provides a dew point sensor inspection system, including: The initial control module 100 is used to keep all probes sealed and closed in the initial state, select one probe to turn on and continuously perform routine on-site dew point detection; The periodic acquisition module 200 is used to activate the dedicated inspection probe and simultaneously acquire dew point detection data corresponding to the conventional detection probe and the inspection probe when the preset inspection cycle is reached. The calculation and comparison module 300 is used to perform difference calculation on two sets of dew point detection data, compare the calculation result with the preset judgment threshold, and obtain the probe working status judgment result. The abnormal switching module 400 is used to activate the backup probe to replace the regular dew point detection if the comparison result exceeds the threshold, and restore the original abnormal detection probe and the inspection probe to the sealed closed state. The cyclical maintenance module 500 is used to trigger the inspection cycle in rounds. Each time, the data is compared with the currently used detection probe through the inspection probe. When all reusable detection probes are abnormal, a fault prompt signal is output.

[0065] Reference Figure 3 This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown, this computer device includes a processor, memory, network interface, and database connected via a bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores operations, computer programs, and the database. The internal memory provides an environment for the operation and execution of the computer programs stored in the non-volatile storage media. The database stores data such as dew point sensor inspection methods. The network interface is used for communication with external terminals via a network connection. When executed by a processor, this computer program implements a dew point sensor inspection method, including the following steps: initially, all probes are kept sealed and closed; one probe is selected and turned on to continuously perform routine on-site dew point detection; upon reaching a preset inspection cycle, a dedicated inspection probe is turned on, and dew point detection data corresponding to the routine detection probe and the inspection probe are collected simultaneously; the difference between the two sets of dew point detection data is calculated, and the calculation result is compared with a preset judgment threshold to obtain the probe working status judgment result; if the comparison result exceeds the threshold, a backup probe is activated to replace the routine dew point detection, and the original abnormal detection probe and the inspection probe are restored to a sealed and closed state; the inspection cycle is triggered in cycles, and each time the data is compared between the inspection probe and the currently used detection probe; when all reusable detection probes are abnormal, a fault indication signal is output.

[0066] One embodiment of this application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements a dew point sensor inspection method, including the following steps: initially, all probes are kept in a sealed and closed state; one probe is selected and turned on to continuously perform routine on-site dew point detection; when a preset inspection cycle is reached, a dedicated inspection probe is turned on to simultaneously collect dew point detection data corresponding to the routine detection probe and the inspection probe; the difference between the two sets of dew point detection data is calculated, and the calculation result is compared with a preset judgment threshold to obtain a probe working status judgment result; if the comparison result exceeds the threshold, a backup probe is activated to replace the routine dew point detection, and the original abnormal detection probe and the inspection probe are restored to a sealed and closed state; the inspection cycle is triggered in cycles, and each time the data is compared between the inspection probe and the currently used detection probe; when all reusable detection probes are abnormal, a fault prompt signal is output.

[0067] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0068] The above description is only a preferred embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural changes made based on the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for inspecting a dew point sensor, characterized in that, include: Initially, all probes are kept sealed and closed. Then, one probe is selected and turned on to continuously perform routine on-site dew point testing. When the preset inspection cycle is reached, the dedicated inspection probe is turned on, and dew point detection data corresponding to the conventional detection probe and the inspection probe are collected simultaneously. The difference between the two sets of dew point detection data is calculated, and the calculation result is compared with the preset judgment threshold to obtain the probe working status judgment result. If the comparison result exceeds the threshold, the backup probe will be activated to replace the regular dew point detection, and the original abnormal detection probe and the inspection probe will be restored to the sealed and closed state. The inspection cycle is triggered in a round-by-round manner. Each time, the data is compared with the currently used detection probe through the inspection probe. When all reusable detection probes are abnormal, a fault prompt signal is output.

2. The inspection method for the dew point sensor according to claim 1, characterized in that, The initial state involves keeping all probes sealed and closed, then selecting one probe to turn on and continuously performing routine on-site dew point testing, including: Read all probe intrinsic parameters, baseline data and five dynamic parameters, construct a multi-dimensional information mapping table, and uniformly mark all probes as sealed and disable the acquisition channel; Based on the constructed information mapping table, a hierarchical screening is carried out. The priority index is calculated by combining hardware qualifications, calibration accuracy and dynamic parameters, and a unique initial candidate detection probe is selected from all probes. Check the sealing status of the candidate probes. If they are qualified, issue an opening command and update the status. If they are not qualified, remove them and re-screen them until a compliant probe is selected and the selection result cannot be changed. Assign independent storage and sampling permissions to the selected probe, activate the dedicated acquisition channel and block the access permissions of other probes, and control it to continuously carry out on-site dew point detection and upload detection data.

3. The inspection method for the dew point sensor according to claim 1, characterized in that, The step of activating the dedicated inspection probe and simultaneously collecting dew point detection data corresponding to the conventional detection probe and the inspection probe when the preset inspection cycle is reached includes: The background runs a timing task and matches the built-in inspection cycle parameters. The cycle is determined by the system timestamp. Once the cycle conditions are met, an inspection trigger event is generated, the dedicated inspection probe is turned on, and the preheating process begins. After the preheating is completed, the sampling sequence numbers of the two sets of probes are reset and cleared to zero. The inspection probes are unlocked, initialized, handshaked, and self-checked to maintain the uninterrupted online acquisition link of the regular detection probes. Send a synchronous trigger enable signal, define the acquisition time window and allocate independent communication time slots, issue a unified sampling scheduling command, control the two probes to carry out sampling in parallel under the same sequence and working conditions and mark the sampling sequence number; Based on the sampling sequence number, two sets of data are paired and encapsulated point-to-point, collected into a dedicated data buffer for inspection, and aligned according to the queue rules. The normalized data is then stored in the corresponding comparison partitions and the regular probe data queue status is restored.

4. The inspection method for the dew point sensor according to claim 1, characterized in that, The step of performing difference calculation on two sets of dew point detection data, comparing the calculation result with a preset judgment threshold, and obtaining the probe working status judgment result includes: Read two sets of dew point detection data with time alignment, complete precise pairing based on synchronization identifier, remove abnormal and invalid data points and define judgment data window, construct standard comparison dataset, and generate signed difference sequence by point-by-point operation; The dataset is segmented and divided into sliding windows. The deviation statistics are calculated segment by segment. The proportion of outliers and the distribution of the sign of the difference are statistically analyzed. The cumulative sum sequence is calculated recursively, and the out-of-bounds points are statistically analyzed. The system retrieves various preset constraint thresholds and judgment limits to conduct comprehensive cross-comparison and verification of segmented deviation, overall deviation, data fluctuation, sign consistency, and cumulative and out-of-bounds situations. Based on the combination and matching relationship of multi-dimensional verification indicators, the normal state of the probe is distinguished from different abnormal types, and a fixed probe working status judgment result is generated.

5. The inspection method for the dew point sensor according to claim 4, characterized in that, The steps of segmenting the dataset and dividing it into sliding windows, calculating the deviation statistics segment by segment, statistically analyzing the proportion of outliers and the distribution of the difference sign within the window, recursively calculating the cumulative sum sequence, and completing the statistics of out-of-bounds points include: Read the signed difference sequence and initialize the cumulative sum benchmark, allocate a dedicated storage area for the sequence, perform validity checks on the difference data, remove invalid data, and complete the classification labeling according to positive and negative types; The cumulative operation is performed separately according to positive and negative categories. The cumulative results of the two types of differences are recorded independently. Simultaneously, the first-level cumulative sum corresponding to the original difference and the second-level cumulative sum sequence corresponding to the absolute value of the difference are generated recursively. The difference data is divided into segments as a whole, and an independent cumulative sum sequence is established for each segment. The deviation statistics are calculated for each segment, the outlier points in each segment are counted and the proportion of local outliers is calculated to form a segmented cumulative sequence. The system iterates through the cumulative sums and segmented subsequences at each level, compares and screens outbound points and abnormal segments, summarizes various statistical indicators, and stores the statistical results and each cumulative sum sequence into their respective independent storage areas.

6. The method for inspecting a dew point sensor according to claim 1, characterized in that, The step of activating a backup probe to replace the regular dew point detection if the comparison result exceeds the threshold, and restoring the original abnormal detection probe and the inspection probe to a sealed and closed state, includes: After determining that the threshold comparison result is abnormal, lock the two types of probe device identifiers and establish a dedicated control link to isolate external signal interference. After reading the status flag bit, issue shutdown commands in sequence and receive corresponding feedback frames. Verify the timing information of the feedback frame to confirm that the instruction is executed effectively, issue a collection termination instruction and shut down the collection channel and communication port, clear the collection buffer, and reissue the compliant shutdown instruction after verifying the communication link through test frame interaction. The system sequentially issues sampling pause, moderate shutdown and complete shutdown control commands in stages, updates the sealing status flag after receiving execution feedback, initiates a bus release request and sets the communication port to the isolated state. The system solidifies the sealing configuration of the two types of probes and completes the status verification. It then disables the relevant control and acquisition interfaces, resets the buffer, address pointer, and counting resources, sends an enable command to the backup probe, and completes the switching and takeover of the routine dew point detection task.

7. The method for inspecting a dew point sensor according to claim 6, characterized in that, After the threshold comparison result is abnormal, the steps of locking the identifiers of the two types of probe devices and establishing a dedicated control link, isolating external signal interference, reading the status flag bit, issuing shutdown commands sequentially level by level, and receiving corresponding feedback frames include: Lock onto the target probe device identifier, establish a dedicated control link and isolate external signal interference, read the status flag bit and complete the validity verification, and create a command interaction waiting record table; Send the first round of shutdown commands to the two types of probes and configure exclusive serial numbers, open the dedicated receiving channel, start the listening mechanism and activate the receiving interrupt to capture feedback frames; After capturing the feedback frame, stop listening and interrupting, record time information, verify the relevant identifiers and codes of the feedback frame, update the record table and send a completion signal after a match is found. After receiving the completion signal, a second round of shutdown command is issued, the interrupt conditions are reconfigured and the listening is restarted, the feedback frame verification is completed, the execution status is marked and the two-level command interaction is confirmed to be complete.

8. The method for inspecting a dew point sensor according to claim 7, characterized in that, The steps of sending the first round of shutdown commands to the two types of probes and configuring exclusive serial numbers, opening a dedicated receiving channel, starting the listening mechanism, and activating the receiving interrupt to capture feedback frames include: Send the first round of shutdown commands with sequence identifiers to both types of probes, open the dedicated receiving channel, save the task context, and define the blocking waiting time boundary; The original execution flow of the program is blocked, the task suspension and semaphore wait functions are called, and the current task is moved into the blocking queue and the semaphore wait queue. Shield system-irrelevant interrupts and low-priority task scheduling, enable dedicated receive interrupts, release semaphores and wake up blocked processes after matching the target feedback frame; After the process resumes running, it verifies the semaphore status and timing information, clears the task suspension flag, closes the dedicated receive channel, destroys the semaphore, and releases the blocked waiting state.

9. The inspection method for the dew point sensor according to claim 1, characterized in that, The step of triggering the inspection cycle in a sequential manner, comparing data between the inspection probe and the currently used detection probe each time, and outputting a fault indication signal when all reusable detection probes are abnormal, includes: Construct a reusable probe circular queue and initialize it according to priority. Independently set up dedicated inspection probes, initialize the inspection state machine and bind the probe identifier. After each round of inspection cycle, establish a fixed pairing relationship between the probes in use and the inspection probes. After the periodic trigger, the inspection state machine is switched, and the paired probe is synchronously driven to complete the unified timing acquisition, cache the acquired data and generate an inspection log, recording the inspection-related identifiers and data summary information; Extract the two detection data streams and perform validity verification and format standardization. Construct a standardized dataset. Use the inspection probe data as a benchmark to perform difference calculation and threshold comparison. Switch the state machine and update the inspection log based on the results. When a probe is found to be abnormal, it is removed from the loop queue, and a subsequent probe is used to take over the detection and the relevant probe status is reset. The inspection process is continuously executed in a loop. When the queue of reusable probes is exhausted, a fault prompt signal is output.

10. A dew point sensor inspection system, characterized in that, include: The initial control module is used to keep all probes sealed and closed in the initial state, select one probe to turn on and continuously perform routine on-site dew point testing; The periodic acquisition module is used to activate the dedicated inspection probe and simultaneously acquire dew point detection data corresponding to the conventional detection probe and the inspection probe when the preset inspection cycle is reached. The calculation and comparison module is used to perform difference calculation on two sets of dew point detection data, compare the calculation result with the preset judgment threshold, and obtain the probe working status judgment result. The abnormal switching module is used to activate the backup probe to replace the regular dew point detection if the comparison result exceeds the threshold, and restore the original abnormal detection probe and the inspection probe to the sealed closed state. The cyclical maintenance module is used to trigger the inspection cycle in rounds. Each time, the data is compared with the currently used detection probe through the inspection probe. When all reusable detection probes are abnormal, a fault prompt signal is output.

Images