Flow switch remote monitoring and early warning system based on internet of things

Abstract

The application discloses a flow switch remote monitoring and early warning system based on an internet of things, and relates to the technical field of industrial internet of things.The application establishes a causal correlation model of a flow switch and an upstream active control device through a causal correlation module, and deduces an expected signal value of the flow switch when a switching event occurs in the upstream device; the heterogenous deviation module is used to calculate the heterogenous deviation amount of an actual signal and an expected signal under the same time stamp; the trusted judgment module is used to statistically evaluate the deviation accumulation intensity in a period, and the deviation accumulation intensity is substituted into a trusted attenuation judgment function to generate a trusted attenuation early warning and push the trusted attenuation early warning to a remote terminal.The application effectively reduces the monitoring false alarm and missed alarm rates, reduces system resource consumption, and can identify the performance deterioration trend of the flow switch in advance.

Description

Technical Field

[0001] This invention relates to the field of industrial Internet of Things (IoT) technology, specifically to an IoT-based remote monitoring and early warning system for flow switches. Background Technology

[0002] As a core flow status detection device in pipeline systems, the reliability of flow switches directly affects the safety and stability of industrial production processes. With the deep penetration of IoT technology in the industrial field, remote monitoring and early warning have become the mainstream method for flow switch status management. However, existing technical solutions still have the following shortcomings: First, the existing monitoring system has not established a causal relationship with the upstream active control equipment, and cannot distinguish whether the signal change is a normal operation response or a fault in the equipment itself, resulting in a high false alarm rate. Secondly, most existing deviation judgment mechanisms use fixed threshold comparisons, without considering the differences in the action characteristics of different upstream devices. At the same time, they do not differentiate the signal deviations caused by devices with different response characteristics, and cannot accurately identify the gradual performance degradation trend of the devices. Third, existing early warning models are mostly based on single abnormality triggering alarms or simple cumulative counting, without combining the dynamic adjustment judgment criteria of the service life of the flow switch, and cannot quantify the decay process of equipment reliability, which is prone to false alarms due to early sporadic abnormalities. Therefore, there is an urgent need for a remote monitoring and early warning system for flow switches based on the Internet of Things. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a remote monitoring and early warning system for flow switches based on the Internet of Things, which solves the problems of false alarms, missed alarms, and high resource consumption in existing flow switch monitoring systems.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a remote monitoring and early warning system for flow switches based on the Internet of Things, comprising: The causal relationship module establishes a causal relationship model between the flow switch and its upstream devices on the cloud platform. When a switching event is detected in the upstream device of the flow switch, the expected signal value of the flow switch corresponding to the switching event is deduced based on the causal relationship model, and an expected signal record is generated using the timestamp of the switching event as an index. The heterogeneous deviation module uses timestamps as indexes to extract actual signal values ​​with the same timestamps as expected signal records from the actual signal data stream of the flow switch collected by the IoT gateway, and calculates the heterogeneous deviation between the actual signal value and the expected signal value at that timestamp. The reliability determination module counts switching events where the signal deviation from the source exceeds the allowable fluctuation range within a preset evaluation period, calculates the cumulative deviation intensity of the evaluation period, and uses the cumulative deviation intensity as input into the reliability attenuation determination function to obtain the current reliability attenuation index. When the index reaches the warning threshold, a reliability attenuation warning is generated for the flow switch, and the warning information, flow switch number, and corresponding evaluation period are pushed to the remote terminal.

[0005] As a further aspect of the present invention, the specific steps for establishing a causal relationship model are as follows: Read the structured design data of the pipeline system, extract the marked equipment and the pipe connections between the equipment, determine the positive direction of each pipe connection along the medium design flow direction, convert each pipe connection into a directed edge, and construct a directed topology graph covering the entire pipe section. Based on the directed topology graph, starting from each flow switch, the reverse medium flow is traced back to the first active control device, and this path is recorded as an upstream influence path of that flow switch; Iterate through all flow switches, extract and store the upstream influence path set corresponding to each flow switch, and for each upstream influence path in the path set, read all operable status bits of the active control device in the path, and at the same time read all observable output status bits of the flow switch at the end of the path. Based on each operable state bit of the active control device in the upstream influence path, the expected output state bit corresponding to the flow switch is determined one by one to obtain the causal state correspondence of the upstream influence path. Repeat this process for all upstream influence paths, and summarize the causal state correspondence of all upstream influence paths to form a causal relationship model.

[0006] As a further aspect of the present invention, the active control device refers to a device in a pipeline system that can respond to external commands and change its own working state, thereby causing a change in the state of the downstream medium.

[0007] As a further aspect of the present invention, the specific rules for capturing switching events occurring upstream of the flow switch are as follows: Connect the status feedback contacts of each active control device upstream of the flow switch to the digital input terminal of the IoT gateway; The gateway operates in a level transition trigger mode. When a level transition is detected at the input terminal, the gateway immediately captures the transition event, reads the state value after the transition, and records the timestamp corresponding to the transition. The gateway encapsulates the status value, timestamp, and corresponding upstream active control device identifier into an event message and uploads it to the cloud platform. After receiving the event message, the cloud platform retrieves the most recently reported status value of the upstream active control device and compares it with the current status value. If they are different, it determines that a switching event has occurred in the upstream active control device, and uses the timestamp in the current event message as the time when the switching event occurred.

[0008] As a further aspect of the present invention, the specific operation of deriving the expected signal value of the flow switch corresponding to the switching event based on the causal relationship model is as follows: When a switching event is detected in one of the upstream active control devices of the flow switch, the current operable status bit of that active control device after the switching is completed is read. Determine the upstream impact path to which the switching event belongs, retrieve the causal state correspondence of the upstream impact path from the causal association model, read the stored expected output state bit from it, and use it as the expected signal value of the flow switch corresponding to the switching event.

[0009] As a further aspect of the present invention, the specific operation for calculating the heterogeneous deviation between the actual signal value and the expected signal value under the timestamp is as follows: When the expected signal value is consistent with the actual signal value, the deviation of the opposite source is recorded as 0. When the expected signal value is inconsistent with the actual signal value, the upstream active control device involved in the switching event that triggered this expected derivation is identified, and the switching event is divided into fast response type deviation and slow response type deviation. A first deviation base value is assigned to fast-response deviations, and a second deviation base value is assigned to slow-response deviations. Both the first and second deviation base values ​​are preset positive real numbers, and the first deviation base value is less than the second deviation base value. Based on the obtained deviation baseline value as input, determine whether there are other switching events within the preset time domain observation window before the timestamp corresponding to the current switching event, and whether the expected signal value corresponding to the switching event is inconsistent with the actual signal value; If it does not exist, the obtained deviation base value is directly used as the heterogeneous deviation amount under that timestamp. If it exists, a correction factor is superimposed on the obtained deviation base value, and the superposition result is used as the heterogeneous deviation amount under that timestamp. The correction factor is a pre-set positive real number.

[0010] As a further aspect of the present invention, the specific rules for classifying switching events into fast-response divergence and slow-response divergence are as follows: Determine the upstream active control device that triggered the current switching event, and read the inherent action duration of the active control device. The inherent action duration refers to the time elapsed from the time the active control device receives the switching command until it completes the medium connection / disconnection or adjustment state change. The inherent action duration is compared with the preset duration: if the inherent action duration is less than the preset duration, the switching event corresponding to the upstream active control device is classified as a fast response type divergence; otherwise, the switching event corresponding to the upstream active control device is classified as a slow response type divergence.

[0011] As a further aspect of the present invention, the specific operation for calculating the cumulative intensity of the deviation during the evaluation period is as follows: Read the set of switching events divided into fast response type deviation and slow response type deviation within the evaluation period, and sum the heterogeneous deviation corresponding to all switching events in each set to obtain the fast response accumulated value and the slow response accumulated value respectively. Based on the set of switching events for fast-response divergence and slow-response divergence within the evaluation period, the timestamps of all switching events are counted and sorted in ascending order to obtain a timestamp sequence. The time interval between adjacent switching events in the timestamp sequence is calculated, and the number of time intervals less than the preset dense judgment interval is counted and used as the number of dense divergences in the evaluation period. In the pre-set table of correspondence between the number of dense divergences and the intensity amplification factor, the intensity amplification factor corresponding to the current time is retrieved by indexing the number of dense divergences, and the accumulated value of the slow response is multiplied by the intensity amplification factor. The resulting product is the cumulative deviation intensity of the evaluation period.

[0012] As a further aspect of the present invention, the specific expression of the credibility decay determination function is as follows: Where D is the credibility decay index, S is the cumulative intensity of deviation, and k is the decay rate coefficient.

[0013] As a further aspect of the present invention, the attenuation rate coefficient k depends on the duration segment to which the current flow switch has been in service for a total period of time, specifically according to the following rule: If the total service time is less than the upper limit threshold of the first section, the flow switch is determined to be in the first section, and the attenuation rate coefficient k is taken as the first preset value k1. If the total service time is greater than or equal to the upper limit threshold of the first segment and less than the upper limit threshold of the second segment, then the flow switch is determined to be in the second segment and the attenuation rate coefficient k is taken as the second preset value k2. If the total service time is greater than or equal to the upper limit threshold of the second section, the flow switch is determined to be in the third section, and the attenuation rate coefficient k is taken as the third preset value k3. Among them, k1, k2, and k3 satisfy k1 < k2 < k3.

[0014] This invention provides a remote monitoring and early warning system for flow switches based on the Internet of Things, which has the following advantages compared with the prior art: (1) By constructing a causal relationship model between the flow switch and the upstream active control equipment, this invention can accurately distinguish whether the change in the flow switch signal is a normal operation response or a fault of the equipment itself, thereby reducing the false alarm rate from the root cause. (2) The present invention adopts a differentiated method for calculating the deviation of different sources. It divides the deviation type according to the inherent action characteristics of the upstream equipment and assigns the corresponding deviation base value. It also combines the historical deviation situation in the time domain observation window with the superimposed correction factor, which can accurately quantify the severity of a single signal deviation, effectively filter out the false deviation caused by time misalignment, and provide a reliable data basis for subsequent fault trend analysis. (3) By combining the cumulative intensity of deviation with the dynamic reliability decay judgment function, and taking into account the time density of deviation events and the service life of the flow switch, this invention can characterize the gradual decay process of equipment reliability, identify the trend of equipment performance deterioration in advance, and avoid false alarms caused by early occasional anomalies. Attached Figure Description

[0015] Figure 1 This is the system principle block diagram of the present invention; Figure 2 This is a flowchart illustrating the steps involved in calculating the heterogeneous deviation under a timestamp in this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] like Figure 1 This invention provides a remote monitoring and early warning system for flow switches based on the Internet of Things; As an embodiment of this application, it includes: The causal relationship module establishes a causal relationship model between the flow switch and its upstream devices on the cloud platform. When a switching event is detected in the upstream device of the flow switch, the expected signal value of the flow switch corresponding to the switching event is deduced based on the causal relationship model, and an expected signal record is generated using the timestamp of the switching event as an index. The heterogeneous deviation module uses timestamps as indexes to extract actual signal values ​​with the same timestamps as expected signal records from the actual signal data stream of the flow switch collected by the IoT gateway, and calculates the heterogeneous deviation between the actual signal value and the expected signal value at that timestamp. The reliability determination module counts switching events where the signal deviation from the source exceeds the allowable fluctuation range within a preset evaluation period, calculates the cumulative deviation intensity of the evaluation period, and uses the cumulative deviation intensity as input into the reliability attenuation determination function to obtain the current reliability attenuation index. When the index reaches the warning threshold, a reliability attenuation warning is generated for the flow switch, and the warning information, flow switch number, and corresponding evaluation period are pushed to the remote terminal.

[0018] As a second embodiment of this application, it is implemented based on the first embodiment, except that this embodiment includes: For an isolated flow switch signal, the computer sees it as just a change in level, which is either 0 or 1. It doesn't know whether this change is caused by normal operation or equipment failure. However, by establishing a causal relationship model between the flow switch and its upstream equipment on the cloud platform, the computer essentially builds a digital causal logic graph for the physical pipeline system in the cloud. Meanwhile, in actual industrial systems, there may be multiple valves and multiple pumps upstream of a flow switch, and the influence relationship between them is combined. For example, the downstream flow switch should only have a flow signal when the upstream valves A and B are open at the same time and the pump C is running. If either condition is missing, there will be no flow. The causal relationship model is a structured way to describe this complex logic with multiple conditions. The specific steps for constructing the causal relationship model are as follows: Read the structured design data of the pipeline system, extract the marked equipment and the pipe connections between each equipment (each pipe connection is associated with the two equipment at its two ends), determine the positive direction of each pipe connection along the medium design flow direction, convert each pipe connection into directed edges, and construct a directed topology graph covering the entire pipe segment; Based on the directed topology graph, starting from each flow switch, the reverse medium flow is traced back to the first active control device, and this path is recorded as an upstream influence path of that flow switch; The active control device refers to a device that can intentionally change the medium in a pipeline from "passable" to "impassable", or from "impassable" to "passable", or change the amount of medium passing through, by receiving control commands or through human operation. Examples include solenoid valves, electric regulating valves, pneumatic shut-off valves, and water pumps. Similarly, starting from the same flow switch, follow the medium flow downstream to the first active control device, and record this path as a downstream influence path of the flow switch; Iterate through all flow switches, extract and store the upstream and downstream impact path sets corresponding to each flow switch; For each upstream influence path in the upstream influence path set, read all operable status bits of the active control devices in that path, and at the same time read all observable output status bits of the flow switch at the end of that path; The operable state bit refers to all the discrete operating states that the active control device itself possesses; Each operable state position corresponds to a deterministic operating state of the active control device. For example, the operable state positions of a two-position switching valve are two discrete values: fully open and fully closed, and the operable state positions of a three-position valve are three discrete values: forward, cut-off, and reverse. The observable output status bit refers to all discrete signal states that the flow switch itself can output; A flow switch is a detection device that can output discrete signals based on the flow state of the medium in a pipeline. Each observable output state bit corresponds to a deterministic output state of the flow switch. For example, the observable output state bit of a baffle-type flow switch has two discrete values: on and off. For each operable state bit of the active control device in the upstream influence path, the expected output state bit corresponding to the flow switch is determined one by one to obtain the causal state correspondence of the upstream influence path. For example, if the active control device is a three-position regulating valve, its operable states are forward, cut-off, and reverse. When it is in the forward state, the medium flows through the flow switch in the forward direction, and the flow switch is expected to output flow. When it is in the reverse state, the medium flows through the flow switch in the reverse direction, and the flow switch is also expected to output flow. When it is in the cut-off state, the medium is blocked, and the flow switch is expected to output no flow. In summary, the two operable states of forward and reverse correspond to the expected output state of flow, and the cut-off state corresponds to the expected output state of no flow. Repeat this process for all upstream influence paths, and summarize the causal status correspondence of all upstream influence paths to form a causal relationship model; When a switching event is detected in the upstream device of the flow switch, the expected signal value of the flow switch corresponding to the switching event is deduced based on the causal relationship model, and an expected signal record is generated using the timestamp of the switching event as an index. For a large IoT system covering hundreds or thousands of flow switches, if the cloud platform needs to continuously poll the data of each sensor and run the comparison algorithm in full, the computational overhead and network bandwidth consumption will be enormous. By adopting the "capture switching event" approach, the system can remain silent and listen most of the time. Only when the state of the upstream device changes will the flow switch derivation and comparison of the path associated with the event be triggered specifically. The specific rules for the upstream switching event of the capture flow switch are as follows: Connect the status feedback contacts of each active control device upstream of the flow switch to the digital input terminal of the IoT gateway; The gateway is configured to capture the transition event when the input level changes, read the current state value after the transition, and record the timestamp of the transition. The gateway encapsulates the status value, timestamp, and corresponding upstream active control device identifier into an event message and uploads it to the cloud platform. After receiving the event message, the cloud platform retrieves the most recently reported status value of the upstream active control device and compares it with the current status value. If they are different, it determines that a switching event has occurred in the upstream active control device, and uses the timestamp in the current event message as the time when the switching event occurred. The specific operation for deriving the expected signal value of the flow switch corresponding to the switching event based on the causal relationship model is as follows: When a switching event is detected in one of the upstream active control devices of the flow switch, the current operable status bit of that active control device after the switching is completed is read. Determine the upstream impact path to which the switching event belongs, retrieve the causal state correspondence of the upstream impact path from the causal association model, read the stored expected output state bit from it, and use it as the expected signal value of the flow switch corresponding to the switching event.

[0019] The heterogeneous deviation module uses timestamps as indexes to extract actual signal values ​​with the same timestamps as expected signal records from the actual signal data stream of the flow switch collected by the IoT gateway, and calculates the heterogeneous deviation between the actual signal value and the expected signal value at that timestamp. In industrial settings, there is an objective physical transition process from the action of upstream equipment to the response of the flow switch. At the same time, there are also differences in the transmission timing when the IoT gateway collects and uploads signals from various devices. If the timestamps are not accurately matched, the old state of the flow switch before the switching event or the new stable state after the switching is completed may be incorrectly compared with the expected value at the moment of switching. Only by comparing at the same timestamp can the expected and actual differences at the moment the switching event is triggered be captured. Even if the difference is due to a delay in the flow switch response or failure to reset in time, this is itself abnormal information that needs to be detected. Meanwhile, subsequent cumulative deviation statistics depend on the currently calculated heterogeneous deviation amount. If the timestamp matching of the extracted actual signal value is not strict, it will cause the calculated deviation amount to be mixed with pseudo-deviations caused by time misalignment. Once these pseudo-deviations enter the cumulative statistics, they will pollute the calculation results of the cumulative deviation intensity, ultimately reducing the accuracy of the reliability decay warning.

[0020] The reliability determination module counts switching events where the signal deviation from a different source exceeds the allowable fluctuation range within a preset evaluation period, and calculates the cumulative deviation intensity for that evaluation period. The heterogeneous deviation is generated with each upstream device switching event. Without time boundary constraints, the deviation events will continue to accumulate over an infinitely long time axis. This accumulation will have a consequence: the small deviations that occurred early and the dense deviations that occurred recently are counted equally, which cannot reflect the latest trend of sensor performance degradation. By setting a preset evaluation period, it is equivalent to defining a fixed-length time window for statistical operations. Each statistical analysis only focuses on deviation events within this window, and events outside the window are not included in the calculation. Furthermore, this window slides forward over time, and the results of each calculation reflect the sensor's performance in the most recent period. The specific operation for calculating the cumulative intensity of the deviation during this evaluation period is as follows: Read the set of switching events divided into fast response type deviation and slow response type deviation within the evaluation period, and sum the heterogeneous deviation corresponding to all switching events in each set to obtain the fast response accumulated value and the slow response accumulated value respectively. Based on the set of switching events for fast-response divergence and slow-response divergence within the evaluation period, the timestamps of all switching events are counted and sorted in ascending order to obtain a timestamp sequence. The time interval between adjacent switching events in the timestamp sequence is calculated, and the number of time intervals less than the preset dense judgment interval is counted and used as the number of dense divergences in the evaluation period. In the pre-set table of correspondence between the number of dense divergences and the intensity amplification factor, the intensity amplification factor corresponding to the current time is retrieved by indexing the number of dense divergences, and the slow response accumulation value is multiplied by the intensity amplification factor. The resulting product is the cumulative deviation intensity of the evaluation period. The cumulative deviation intensity is used as input into the credibility decay judgment function to obtain the current credibility decay index. When the index reaches the warning threshold, a credibility decay warning is generated for the flow switch, and the warning information, flow switch number and corresponding evaluation period are pushed to the remote terminal. The specific expression for the credibility decay determination function is as follows: ; Where D is the confidence decay index, with a value range of [0,1]; S is the cumulative bias intensity; and k is the decay rate coefficient. The value of k depends on the duration segment to which the current flow switch has been in service for the total duration, and the specific rules are as follows: If the total service time is less than the upper limit threshold of the first section, the flow switch is determined to be in the first section, and the attenuation rate coefficient k is taken as the first preset value k1. If the total service time is greater than or equal to the upper limit threshold of the first segment and less than the upper limit threshold of the second segment, then the flow switch is determined to be in the second segment and the attenuation rate coefficient k is taken as the second preset value k2. If the total service time is greater than or equal to the upper limit threshold of the second section, the flow switch is determined to be in the third section, and the attenuation rate coefficient k is taken as the third preset value k3. Among them, k1, k2, and k3 satisfy k1 < k2 < k3.

[0021] As a third embodiment of this application, this embodiment further discloses a method for calculating the heterogeneous deviation under a timestamp, based on embodiments one and two. Figure 2 As shown, the specific content includes: When the expected signal value is consistent with the actual signal value, the deviation of the opposite source is recorded as 0. When the expected signal value is inconsistent with the actual signal value, the upstream active control device involved in the switching event that triggered this expected derivation is identified, and the switching event is divided into fast response type deviation and slow response type deviation. The specific rules for classifying switching events into fast-response divergences and slow-response divergences are as follows: Determine the upstream active control device that triggered the current switching event, and read the inherent action duration of the active control device. The inherent action duration refers to the time elapsed from the time the active control device receives the switching command until it completes the medium connection / disconnection or adjustment state change. The inherent action duration is compared with the preset duration: if the inherent action duration is less than the preset duration, the switching event corresponding to the upstream active control device is classified as a fast response type divergence; otherwise, the switching event corresponding to the upstream active control device is classified as a slow response type divergence. A first deviation base value is assigned to fast-response deviations, and a second deviation base value is assigned to slow-response deviations. Both the first and second deviation base values ​​are preset positive real numbers, and the first deviation base value is less than the second deviation base value. Based on the obtained deviation baseline value as input, determine whether there are other switching events within the preset time domain observation window before the timestamp corresponding to the current switching event, and whether the expected signal value corresponding to the switching event is inconsistent with the actual signal value; If it does not exist, the obtained deviation base value is directly used as the heterogeneous deviation amount under that timestamp. If it exists, a correction factor is superimposed on the obtained deviation base value, and the superposition result is used as the heterogeneous deviation amount under that timestamp. The correction factor is a pre-set positive real number.

[0022] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0023] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A remote monitoring and early warning system for flow switches based on the Internet of Things, characterized in that, include: The causal relationship module establishes a causal relationship model between the flow switch and its upstream devices on the cloud platform. When a switching event is detected in the upstream device of the flow switch, the expected signal value of the flow switch corresponding to the switching event is deduced based on the causal relationship model, and an expected signal record is generated using the timestamp of the switching event as an index. The heterogeneous deviation module uses timestamps as indexes to extract actual signal values ​​with the same timestamps as expected signal records from the actual signal data stream of the flow switch collected by the IoT gateway, and calculates the heterogeneous deviation between the actual signal value and the expected signal value at that timestamp. The reliability determination module counts switching events where the signal deviation from the source exceeds the allowable fluctuation range within a preset evaluation period, calculates the cumulative deviation intensity of the evaluation period, and uses the cumulative deviation intensity as input into the reliability attenuation determination function to obtain the current reliability attenuation index. When the index reaches the warning threshold, a reliability attenuation warning is generated for the flow switch, and the warning information, flow switch number, and corresponding evaluation period are pushed to the remote terminal.

2. The IoT-based remote monitoring and early warning system for flow switches according to claim 1, characterized in that, The specific steps for establishing a causal relationship model are as follows: Read the structured design data of the pipeline system, extract the marked equipment and the pipe connections between the equipment, determine the positive direction of each pipe connection along the medium design flow direction, convert each pipe connection into a directed edge, and construct a directed topology graph covering the entire pipe section. Based on the directed topology graph, starting from each flow switch, the reverse medium flow is traced back to the first active control device, and this path is recorded as an upstream influence path of that flow switch; Iterate through all flow switches, extract and store the upstream influence path set corresponding to each flow switch, and for each upstream influence path in the path set, read all operable status bits of the active control device in the path, and at the same time read all observable output status bits of the flow switch at the end of the path. Based on each operable state bit of the active control device in the upstream influence path, the expected output state bit corresponding to the flow switch is determined one by one to obtain the causal state correspondence of the upstream influence path. Repeat this process for all upstream influence paths, and summarize the causal state correspondence of all upstream influence paths to form a causal relationship model.

3. The IoT-based remote monitoring and early warning system for flow switches according to claim 2, characterized in that, The active control device refers to a device in a pipeline system that can respond to external commands and change its own working state, thereby causing a change in the state of the downstream medium.

4. The IoT-based remote monitoring and early warning system for flow switches according to claim 1, characterized in that, The specific rules for capturing switching events upstream of the flow switch are as follows: Connect the status feedback contacts of each active control device upstream of the flow switch to the digital input terminal of the IoT gateway; The gateway operates in a level transition trigger mode. When a level transition is detected at the input terminal, the gateway immediately captures the transition event, reads the state value after the transition, and records the timestamp corresponding to the transition. The gateway encapsulates the status value, timestamp, and corresponding upstream active control device identifier into an event message and uploads it to the cloud platform. After receiving the event message, the cloud platform retrieves the most recently reported status value of the upstream active control device and compares it with the current status value. If they are different, it determines that a switching event has occurred in the upstream active control device, and uses the timestamp in the current event message as the time when the switching event occurred.

5. The IoT-based remote monitoring and early warning system for flow switches according to claim 1, characterized in that, Based on the causal relationship model, the specific operation for deriving the expected signal value of the flow switch corresponding to this switching event is as follows: When a switching event is detected in one of the upstream active control devices of the flow switch, the current operable status bit of that active control device after the switching is completed is read. Determine the upstream impact path to which the switching event belongs, retrieve the causal state correspondence of the upstream impact path from the causal association model, read the stored expected output state bit from it, and use it as the expected signal value of the flow switch corresponding to the switching event.

6. The IoT-based remote monitoring and early warning system for flow switches according to claim 1, characterized in that, The specific steps for calculating the heterogeneous deviation between the actual signal value and the expected signal value at the timestamp are as follows: When the expected signal value is consistent with the actual signal value, the deviation of the opposite source is recorded as 0. When the expected signal value is inconsistent with the actual signal value, the upstream active control device involved in the switching event that triggered this expected derivation is identified, and the switching event is divided into fast response type deviation and slow response type deviation. A first deviation base value is assigned to fast-response deviations, and a second deviation base value is assigned to slow-response deviations. Both the first and second deviation base values ​​are preset positive real numbers, and the first deviation base value is less than the second deviation base value. Based on the obtained deviation baseline value as input, determine whether there are other switching events within the preset time domain observation window before the timestamp corresponding to the current switching event, and whether the expected signal value corresponding to the switching event is inconsistent with the actual signal value; If it does not exist, the obtained deviation base value is directly used as the heterogeneous deviation amount under that timestamp. If it exists, a correction factor is superimposed on the obtained deviation base value, and the superposition result is used as the heterogeneous deviation amount under that timestamp. The correction factor is a pre-set positive real number.

7. A remote monitoring and early warning system for flow switches based on the Internet of Things according to claim 6, characterized in that, The specific rules for classifying switching events into fast-response divergences and slow-response divergences are as follows: Determine the upstream active control device that triggered the current switching event, and read the inherent action duration of the active control device. The inherent action duration refers to the time elapsed from the time the active control device receives the switching command until it completes the medium connection / disconnection or adjustment state change. The inherent action duration is compared with the preset duration: if the inherent action duration is less than the preset duration, the switching event corresponding to the upstream active control device is classified as a fast response type divergence; otherwise, the switching event corresponding to the upstream active control device is classified as a slow response type divergence.

8. The IoT-based remote monitoring and early warning system for flow switches according to claim 1, characterized in that, The specific steps for calculating the cumulative intensity of deviation for this assessment period are as follows: Read the set of switching events divided into fast response type deviation and slow response type deviation within the evaluation period, and sum the heterogeneous deviation corresponding to all switching events in each set to obtain the fast response accumulated value and the slow response accumulated value respectively. Based on the set of switching events for fast-response divergence and slow-response divergence within the evaluation period, the timestamps of all switching events are counted and sorted in ascending order to obtain a timestamp sequence. The time interval between adjacent switching events in the timestamp sequence is calculated, and the number of time intervals less than the preset dense judgment interval is counted and used as the number of dense divergences in the evaluation period. In the pre-set table of correspondence between the number of dense divergences and the intensity amplification factor, the intensity amplification factor corresponding to the current time is retrieved by indexing the number of dense divergences, and the accumulated value of the slow response is multiplied by the intensity amplification factor. The resulting product is the cumulative deviation intensity of the evaluation period.

9. A remote monitoring and early warning system for flow switches based on the Internet of Things according to claim 1, characterized in that, The specific expression for the credibility decay determination function is as follows: Where D is the credibility decay index, S is the cumulative intensity of deviation, and k is the decay rate coefficient.

10. A remote monitoring and early warning system for flow switches based on the Internet of Things according to claim 9, characterized in that, The attenuation rate coefficient k depends on the duration segment to which the current flow switch has been in service for the total duration, and the specific rule is as follows: If the total service time is less than the upper limit threshold of the first section, the flow switch is determined to be in the first section, and the attenuation rate coefficient k is taken as the first preset value k1. If the total service time is greater than or equal to the upper limit threshold of the first segment and less than the upper limit threshold of the second segment, then the flow switch is determined to be in the second segment and the attenuation rate coefficient k is taken as the second preset value k2. If the total service time is greater than or equal to the upper limit threshold of the second section, the flow switch is determined to be in the third section, and the attenuation rate coefficient k is taken as the third preset value k3. Among them, k1, k2, and k3 satisfy k1 < k2 < k3.

Images