A condition monitoring system for road anti-collision guardrail

Abstract

The present invention relates to the field of crash barrier monitoring technology, specifically a state monitoring system for road crash barriers, the system comprising: a data acquisition module, which collects sensor data of the crash barrier, reads the impact force, impact frequency and position information of each sensor, records the numerical value and time stamp of the impact force, impact frequency and position information, verifies the data, excludes abnormal values, integrates all the impact force, impact frequency and position information data, and generates a unified data format. In the present invention, the accuracy of the data and the speed of event response are improved by data acquisition and real-time analysis of the impact information of the road crash barrier. And through data verification, abnormal values ​​can be excluded to ensure that only reliable data is used for subsequent analysis, thereby enhancing the accurate judgment of the collision event. In addition, the dynamic update function of automatically adjusting the threshold and updating the rules improves the adaptability and effectiveness of preventive measures based on historical data analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of anti-collision guardrail monitoring, and in particular to a state monitoring system for road anti-collision guardrails. Background Art

[0002] The field of anti-collision guardrail monitoring technology focuses on real-time monitoring and evaluation of the status and safety of guardrails in road traffic, including the use of sensors, cameras and other monitoring equipment to detect the structural integrity, damage level and performance of guardrails in traffic accidents.

[0003] Existing technologies rely on periodic manual inspection and data processing, which limits the system's ability to respond quickly to emergencies. Especially on high-traffic roads, slow response speeds and lack of timely updated data processing rules may lead to poor monitoring results and increase the risk of accidents caused by monitoring delays. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a state monitoring system for road anti-collision guardrails.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: A state monitoring system for road anti-collision guardrail comprises:

[0006] The data acquisition module collects sensor data of the anti-collision guardrail, reads the impact force, impact frequency and position information of each sensor, records the value and time stamp of the impact force, impact frequency and position information, verifies the data, excludes abnormal values, integrates all the impact force, impact frequency and position information data, generates a unified data format, and obtains real-time sensor data;

[0007] A real-time transmission module receives the real-time sensor data, arranges the real-time sensor data in chronological order, caches the data stream, monitors the integrity and order of the data, detects whether there is data loss or duplication, marks the data, and generates a real-time data stream;

[0008] An event recognition module analyzes the impact force, impact frequency and location information in the real-time data stream, compares the value of each parameter with a set threshold, determines whether the impact force exceeds the threshold, checks whether the impact frequency occurs multiple times in a short period of time, identifies the impact event that occurred, records the time, intensity and location of the event, and generates impact event information;

[0009] A rule execution module reads the collision event information, checks the collision force, frequency and position parameters according to predefined rules, determines whether the parameters meet the rule conditions, generates an instruction to send an alarm if the collision force exceeds the target value and the frequency is higher than the set number of times, and generates a notification maintenance instruction if the position parameter is located in a critical section, executes sending an alarm, notifying maintenance and dispatching a maintenance team, and generates a response instruction set;

[0010] The dynamic update module counts past impact events according to the response instruction set and historical data, analyzes the changing trends of impact force, frequency and position parameters, adjusts the thresholds and conditions in the predefined rules, updates the rules in the rule library, records the modification time and reasons of the rules, and obtains the updated rule set.

[0011] Preferably, the steps of acquiring the real-time sensor data are:

[0012] Collect data from each sensor on the anti-collision guardrail, including impact force, impact frequency and location information, record the data value and timestamp, integrate and generate impact data in a unified format to obtain an impact data set;

[0013] Verifying and eliminating outliers on the collision data set, using a cleaning algorithm to process abnormal data points, ensuring data quality and integrity, and generating a cleaned collision data set;

[0014] The cleaned impact data set is integrated to aggregate all impact force, impact frequency and position information data using the formula:

[0015]

[0016] Calculate the weighted impact force index to obtain real-time sensor data, where: Represents the impact force, is the impact force weight, represents the impact frequency, is the frequency weight, Represents location information, is the position weight.

[0017] Preferably, the steps of acquiring the real-time data stream are:

[0018] Receiving the real-time sensor data, arranging the data in chronological order, maintaining the continuity of the timeline and the timing correctness of the data, and generating a time-ordered data set;

[0019] caching the time-ordered data set, monitoring the integrity and order of the data stream in real time, checking whether the data is lost or duplicated, marking abnormal data points, and generating a monitored data stream;

[0020] From the monitored data stream, all repeated or lost data events are detected and marked, and the repeated data are cleared to obtain an optimized real-time data stream.

[0021] Preferably, the steps of identifying the collision event are:

[0022] Analyzing the optimized real-time data stream, comparing the impact force, impact frequency and position information of each data point with a preset threshold, marking the data points exceeding the threshold, and obtaining an over-threshold data set;

[0023] From the above-threshold data set, the impact events with a frequency exceeding the threshold in a short period of time are identified using the formula:

[0024]

[0025] Calculate which events meet the dual conditions of impact force and frequency, and generate a preliminary impact event identification set, where Represents the impact force, is the impact force threshold, represents the impact frequency, is the frequency threshold;

[0026] The preliminary impact event identification set is refined, the location information is analyzed, and it is verified which impact event locations are key areas. The time, intensity and location information are integrated to complete the identification of the impact event.

[0027] Preferably, the steps of determining whether the impact force exceeds the threshold are:

[0028] Analyzing the data in the real-time data stream, including the impact force, impact frequency and position information, and comparing them with preset thresholds to determine whether each parameter exceeds the threshold, and generating a threshold comparison data set;

[0029] The threshold comparison data set is analyzed using the formula:

[0030]

[0031] Calculate the deviation of each data point from the threshold, where Represents the impact force, is the impact force threshold, represents the impact frequency, For the frequency threshold, a deviation calculation result set is generated;

[0032] From the deviation calculation result set, the data point with the largest deviation is selected to determine the data point with the largest deviation as a high-risk candidate for the impact event, and the impact force, frequency and position information of the data point with the largest deviation is recorded and sorted to obtain the impact event threshold comparison result.

[0033] Preferably, the steps of obtaining the response instruction set are:

[0034] Read the impact event information, analyze the impact force, impact frequency and position parameters, determine whether the impact force, impact frequency and position parameters meet predefined rule conditions, and obtain a rule matching result set;

[0035] According to the data in the rule matching result set, determine which collision events meet the conditions of alarm or maintenance, such as the collision force exceeds the target value and the frequency is higher than the set number of times or the location is located in a critical section, and generate an early warning and maintenance decision set;

[0036] Based on the warning and maintenance decision set, the formula is adopted:

[0037]

[0038] Calculate the emergency index of each decision, guide the decision of sending alarms, notifying maintenance or dispatching repair teams, and generate a response instruction set; Represents the degree of compliance with the alarm conditions for each event, is the alarm weight, Represents the degree of compliance with maintenance conditions. To maintain weight.

[0039] Preferably, the step of acquiring the update rule set is:

[0040] Count past impact events, analyze the changing trends of impact force, impact frequency and position parameters, calculate the average change and standard deviation of the parameters, and obtain a statistical analysis result set;

[0041] Based on the statistical analysis result set, the need for rule adjustment is identified using the formula:

[0042]

[0043] Calculation rule parameters, where Indicates the current rule parameters, To adjust the intensity factor, and are the average changes in position and frequency, respectively, and are the standard deviations of position and frequency, respectively, is the standard deviation adjustment coefficient and generates a new rule candidate set.

[0044] Preferably, the step of acquiring the updated rule set further includes: evaluating the effect of the new rule candidate set, verifying the accuracy and effect of the new rules through simulation data, recording the time and reason of rule modification, and obtaining the updated rule set.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are:

[0046] In the present invention, the accuracy of data and the speed of event response are improved by data collection and real-time analysis of collision information of road anti-collision guardrails. And through data verification, outliers can be excluded to ensure that only reliable data is used for subsequent analysis, thereby enhancing the accurate determination of collision events. In addition, the dynamic update function of automatically adjusting thresholds and updating rules improves the adaptability and effectiveness of preventive measures based on historical data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0049] See also Figure 1 The present invention provides a technical solution: a state monitoring system for a road anti-collision guardrail comprises:

[0050] The data acquisition module collects sensor data of the anti-collision guardrail, reads the impact force, impact frequency and position information of each sensor, records the value and time stamp of the impact force, impact frequency and position information, verifies the data, excludes abnormal values, integrates all the impact force, impact frequency and position information data, generates a unified data format, and obtains real-time sensor data;

[0051] The real-time transmission module receives real-time sensor data, arranges the real-time sensor data in chronological order, caches the data stream, monitors the integrity and order of the data, detects whether there is data loss or duplication, marks the data, and generates a real-time data stream;

[0052] The event recognition module analyzes the impact force, impact frequency and location information in the real-time data stream, compares the value of each parameter with the set threshold, determines whether the impact force exceeds the threshold, checks whether the impact frequency occurs multiple times in a short period of time, identifies the impact event, records the time, intensity and location of the event, and generates impact event information;

[0053] The rule execution module reads the collision event information and checks the collision force, frequency and position parameters according to the predefined rules to determine whether the parameters meet the rule conditions. If the collision force exceeds the target value and the frequency is higher than the set number of times, an instruction to send an alarm is generated. If the position parameter is located in a critical section, a maintenance notification instruction is generated to send an alarm, notify maintenance and dispatch the maintenance team, and generate a response instruction set.

[0054] The dynamic update module counts past impact events based on the response instruction set and historical data, analyzes the changing trends of impact force, frequency and position parameters, adjusts the thresholds and conditions in the predefined rules, updates the rules in the rule base, records the modification time and reasons of the rules, and obtains the updated rule set.

[0055] The steps to obtain real-time sensor data are:

[0056] Collect data from each sensor on the anti-collision guardrail, including impact force, impact frequency and location information, record the data value and timestamp, integrate and generate impact data in a unified format to obtain an impact data set;

[0057] Verify the impact data set and eliminate outliers, use the cleaning algorithm to process abnormal data points, ensure data quality and integrity, and generate a cleaned impact data set;

[0058] The cleaned impact data set is integrated to aggregate all impact force, impact frequency and location information data using the formula:

[0059]

[0060] Calculate the weighted impact force index to obtain real-time sensor data, where: Represents the impact force, is the impact force weight, represents the impact frequency, is the frequency weight, Represents location information, is the position weight.

[0061] Specifically, the data acquisition module is started, and the sensors of the anti-collision guardrail are initialized and calibrated to ensure that the impact force, impact frequency and position information can be accurately collected in actual operation. The sensor monitors the force, frequency and specific location of the impact event in real time, and converts the analog signal into a digital signal through digital signal processing technology for subsequent data processing and analysis. The data integration process includes formatting and time-stamping the collected indicators to ensure the timing correctness and integrity of the data. The recorded data is preliminarily screened to remove those obviously abnormal values, such as non-actual impact data caused by sensor failure or external interference. The impact data set generated after integration is the result of preliminary cleaning and screening of the original data to ensure the accuracy of subsequent analysis.

[0062] After completing the initial data collection, the data cleaning algorithm is used to further process the impact data set. Statistical analysis methods are used to identify and eliminate outliers in the data. These outliers may be caused by sensor failure, environmental interference or data transmission errors. Data cleaning includes verifying the integrity and consistency of the data to ensure that each data meets the preset quality standards. The cleaned data is more representative and reliable. The data is standardized to unify the data recording standards for different sensors and under different conditions. The cleaned impact data set provides a high-quality data foundation for subsequent analysis. These data will be used to generate real-time data streams to monitor and analyze impact events.

[0063] formula The calculation process:

[0064] Represents the comprehensive impact force index, , and Represent the impact force, impact frequency and position information of the i-th impact respectively, and the weight , and Assign importance based on impact force, frequency, and location. For example, the impact force weight may be higher based on the severity of the impact, the impact frequency weight may be adjusted based on accident-prone areas, and the location weight may be set based on traffic volume and accident rate;

[0065] There are three impact data: Newton, times / minute, (high-risk position), the weights are set to , , , the calculation process is as follows:

[0066]

[0067] After performing the same calculation for each impact, all Add up the values The results show that the comprehensive impact force index is 90.7, which represents the severity of the impact event and means that an immediate response is needed to prevent further damage.

[0068] The steps to obtain real-time data stream are:

[0069] Receive real-time sensor data, arrange the data in chronological order, maintain the continuity of the timeline and the temporal correctness of the data, and generate a time-ordered data set;

[0070] Cache the time-ordered data set, monitor the integrity and order of the data stream in real time, check whether the data is lost or duplicated, mark abnormal data points, and generate monitored data streams;

[0071] From the monitored data stream, detect and mark all duplicate or lost data events, remove duplicate data, and obtain an optimized real-time data stream.

[0072] Specifically, receive real-time sensor data, arrange these data in chronological order to maintain the continuity of the timeline and the timing correctness of the data, and generate a time-sorted data set. During the receiving process, the sensor first transmits the collected data through wireless signals, including impact force, impact frequency and location information. Each data point contains a timestamp. The system uses these timestamps to sort the data. The sorting algorithm processes the data according to the order of the timestamps to ensure that the data stream enters the system's internal buffer area in chronological order. The correct arrangement of the time sequence is crucial for subsequent data analysis because it ensures the clarity of the logical order and causal relationship of events. This process ensures that each data point can be processed in the correct time frame, thereby generating a time-sorted data set;

[0073] Cache the time-ordered data sets, monitor the integrity and order of the data stream in real time, ensure that there is no data loss or duplication, mark abnormal data points, and generate a monitored data stream. During the caching process, the data first enters a temporary storage area, where a monitoring system checks the integrity and order of each data packet in real time. The monitoring system uses hash functions and timestamps to verify whether the data packet is complete and undamaged, and ensure that the data has not been tampered with or received repeatedly. In addition, the system also detects abnormal data points in the data stream, such as sudden jumps in data values ​​or unexpected data patterns, so that each data point is verified and marked, thereby generating an error-free monitored data stream;

[0074] From the monitored data stream, detect and mark all duplicate or missing data events, clear the duplicate data, and obtain the optimized real-time data stream. During the optimization process, the data management system re-checks the monitored data stream, identifies and processes duplicate data events and data loss events in the data stream. For duplicate data, the system identifies and deletes duplicates by comparing the timestamps and other metadata of the data points. For missing data, the system uses interpolation methods or prediction models based on historical data to complete the missing data points to ensure the quality and availability of the data stream. The optimized real-time data stream is obtained after these processes, which ensures the accuracy and practicality of the data.

[0075] The steps to identify a collision event are:

[0076] Analyze the optimized real-time data stream, compare the impact force, impact frequency and location information of each data point with the preset threshold, mark the data points exceeding the threshold, and obtain the super-threshold data set;

[0077] From the over-threshold data set, we identify the impact events whose frequency exceeds the threshold in a short period of time, using the formula:

[0078]

[0079] Calculate which events meet the dual conditions of impact force and frequency, and generate a preliminary impact event identification set, where Represents the impact force, is the impact force threshold, represents the impact frequency, is the frequency threshold;

[0080] The preliminary impact event identification set is refined, the location information is analyzed, and it is verified which impact event locations are key areas. The time, intensity and location information are integrated to complete the identification of the impact event.

[0081] Specifically, the optimized real-time data stream is continuously monitored by the data acquisition module, and the data includes impact force, impact frequency and location information. These data are updated in real time and saved in the database. Each data point has a clear timestamp and source identification. By comparing with the preset threshold, the system automatically marks those data points that exceed the threshold. For example, if the impact force exceeds 100 Newtons or the impact frequency exceeds 5 times per minute, these data points are regarded as potential danger signals exceeding the threshold. The system not only records these data points exceeding the threshold, but also performs preliminary classification to prepare for subsequent analysis. This process ensures real-time updating and monitoring of the data, while also optimizing the processing efficiency of the data stream, ensuring that all important information can be captured and utilized in a timely manner to obtain an ultra-threshold data set.

[0082] formula The calculation process:

[0083] Indicates the impact force, is the impact force threshold, set to 100 Newtons, represents the impact frequency, is the frequency threshold, set to 5 times / minute. For example, if the impact force of an event is 120 Newtons and the frequency is 6 times / minute, then is true, indicating that this is a valid impact event. This judgment process is performed on each data point in the real-time data stream. Each point is checked to see if it satisfies both conditions. If so, the data point is added to the preliminary impact event identification set. This step is critical because it is directly related to whether the impact event can be correctly identified and responded to, ensuring that the system's response is timely and accurate.

[0084] Continuing the analysis from the initial impact event identification, the location information of these data points is refined, for example, the exact geographic location of each impact event is confirmed through the GIS system, and the time information is synchronized and calibrated to ensure that the timestamp of each event is accurate to the second. By combining the location and time information, the system can track the specific section and time of the impact event, which is very critical for emergency response and subsequent maintenance, and completes the identification of the impact event.

[0085] The steps to determine whether the impact force exceeds the threshold are:

[0086] Analyze the data in the real-time data stream, including impact force, impact frequency and location information, compare with preset thresholds, determine whether each parameter exceeds the threshold, and generate a threshold comparison data set;

[0087] The threshold comparison data set is analyzed using the formula:

[0088]

[0089] Calculate the deviation of each data point from the threshold, where Represents the impact force, is the impact force threshold, represents the impact frequency, For the frequency threshold, a deviation calculation result set is generated;

[0090] From the deviation calculation result set, the data point with the largest deviation is selected to determine the data point with the largest deviation as a high-risk candidate for the impact event. The impact force, frequency and position information of the data point with the largest deviation is recorded and sorted to obtain the impact event threshold comparison result.

[0091] Specifically, the formula The calculation process:

[0092] The impact force of a data point The impact force threshold is 150 Newtons. is 100 Newtons, the impact frequency The frequency threshold is 10 times / minute. 5 times / minute.

[0093] First calculate the difference between the impact force and the threshold:

[0094]

[0095] Then calculate the square root of the difference between the frequency and the threshold:

[0096]

[0097] Adding these two values ​​gives:

[0098]

[0099] This value represents the total deviation between the data point and the threshold. A larger deviation value means that the event is a candidate for a high-risk impact event, which provides a basis for subsequent processing and generates a deviation calculation result set.

[0100] The data points with the largest deviations are selected from the deviation calculation result set. These data points are considered to be candidates for high-risk impact events. For example, all data points with deviation values ​​exceeding 50 are selected, and the specific impact force, frequency, and location information of these data points are recorded. This information is organized into a report with a unified format. This can help the maintenance team quickly locate and respond to potential high-risk impact events, ensure the timeliness and effectiveness of response measures, and obtain impact event threshold comparison results.

[0101] The steps to obtain the response instruction set are:

[0102] Read the impact event information, analyze the impact force, impact frequency and position parameters, determine whether the impact force, impact frequency and position parameters meet the predefined rule conditions, and obtain the rule matching result set;

[0103] According to the data in the rule matching result set, determine which collision events meet the conditions for alarm or maintenance, such as the collision force exceeds the target value and the frequency is higher than the set number of times or the location is located in a critical section, and generate an early warning and maintenance decision set;

[0104] Based on the early warning and maintenance decision set, the formula is adopted:

[0105]

[0106] Calculate the emergency index of each decision, guide the decision of sending alarms, notifying maintenance or dispatching repair teams, and generate a response instruction set; Represents the degree of compliance with the alarm conditions for each event, is the alarm weight, Represents the degree of compliance with maintenance conditions. To maintain weight.

[0107] Specifically, the acquisition of the rule matching result set starts with the analysis of the impact event information, in which the impact force, impact frequency and position parameters involved are obtained through the data stream analysis of the real-time monitoring system. The data stream is continuously monitored by sensors. Whenever an impact event occurs, the relevant data is recorded in real time and transmitted to the central database. After the central database uses a data cleaning algorithm to remove noise, these preliminarily screened data will be standardized to meet the format requirements of the preset rules. Next, the system will compare these standardized data with the predefined thresholds of impact force, frequency and position parameters to ensure real-time updating and accuracy of the data. Through this process, the data is not only verified for its authenticity, but also classified into categories that meet the warning and maintenance conditions. Finally, this process generates a rule matching result set.

[0108] The warning and maintenance decision set is obtained by further decision processing based on the rule matching result set. By analyzing whether the impact force and impact frequency of each impact event exceed the preset target value, and whether the impact location is located in a critical section, these data points are first grouped, and each group of data is analyzed independently to evaluate whether it meets the conditions for sending an alarm or performing maintenance. During the decision-making process, the system will call the historical data in the database in real time for comparative analysis to ensure the accuracy and timeliness of the decision. Through these steps, the system can dynamically adjust the response strategy to ensure the most effective response in the shortest time, and finally generate a warning and maintenance decision set.

[0109] formula The calculation process:

[0110] The urgency index, which represents the response instruction set, is obtained by comprehensively evaluating the severity of the impact event and the urgency of its needs. Indicates the compliance degree of the alarm condition of each event. The specific acquisition method is to compare and analyze whether the impact force of the event exceeds the set threshold. Indicates the degree of compliance with maintenance conditions for each event, and determines whether it is located in a critical section based on location information. and is a weighting factor, which is adjusted based on the response effects of past events and engineer evaluation.

[0111] There are three events, among which , corresponding to ,and ;

[0112] The calculation formula is:

[0113]

[0114] The results showed that the emergency index was 1.86, which represents a comprehensive response index based on the current event severity and maintenance needs to guide actual alarm and maintenance decisions;

[0115] The following thresholds can be set:

[0116] Low-level response: When Below 0.5, only routine monitoring or follow-up may be required.

[0117] Medium response: When When the value is between 0.5 and 1.5, the maintenance team needs to be notified immediately to conduct on-site assessment, but actual operation may not be performed immediately.

[0118] High-level response: When When it exceeds 1.5, it indicates that the emergency situation is serious and emergency response procedures need to be initiated immediately, such as sending an alarm, dispatching a maintenance team, etc.

[0119] In practical applications, the threshold should be adjusted dynamically based on historical data, the severity of the impact event, and the criticality of the location. If the value does not usually lead to serious consequences, the threshold can be adjusted higher to avoid over-response. On the contrary, if small-scale events have led to major impacts, the threshold can be lowered to increase the sensitivity of the warning.

[0120] The steps to obtain the updated rule set are:

[0121] Count past impact events, analyze the changing trends of impact force, impact frequency and position parameters, calculate the average change and standard deviation of the parameters, and obtain a statistical analysis result set;

[0122] Based on the statistical analysis result set, identify the need for rule adjustment using the formula:

[0123]

[0124] Calculation rule parameters, where Indicates the current rule parameters, To adjust the intensity factor, and are the average changes in position and frequency, respectively, and are the standard deviations of position and frequency, respectively, is the standard deviation adjustment coefficient, generating a new rule candidate set;

[0125] Evaluate the effectiveness of the new rule candidate set, verify the accuracy and effectiveness of the new rules through simulation data, record the time and reason for rule modification, and obtain the updated rule set.

[0126] Specifically, we statistically analyzed past impact events, extracted the strength, frequency, and location information of each impact from the data, organized these data, and calculated the mean and standard deviation of each parameter. This involves a large amount of data set organization and mathematical calculations. First, we need to establish a data processing process, which includes three steps: data cleaning, data classification, and data analysis. Data cleaning mainly removes incomplete or erroneous data records, and data classification is based on geographic location and impact intensity. Finally, in the data analysis stage, statistical methods are used to calculate the mean and standard deviation. In order to ensure the accuracy of the calculation, each step needs to be supported by software tools, such as using SQL for data query and preliminary processing, and then using the Pandas library in Python for more in-depth statistical analysis. Through these detailed steps, we finally obtained a reliable set of statistical analysis results.

[0127] formula: The calculation process:

[0128] is the current rule parameter value, assuming it is 1.5, is the adjustment intensity factor, set to 0.05, indicating the fine-tuning amplitude of each adjustment; and are the average changes in position and frequency, respectively. These values ​​are calculated based on the data changes over a period of time. Assume , ; and is the standard deviation of position and frequency, calculated as , ; is the standard deviation adjustment factor, which is set to 0.5 to adjust the impact of the standard deviation. The calculation process is:

[0129]

[0130] This result indicates that the new rule parameters are slightly lowered according to the changes in the impact data, meaning that the response strategy will become slightly more relaxed to adapt to the changing environmental conditions.

[0131] Evaluate the effectiveness of the new rule candidate set, conduct simulation tests on the new rule parameters, and verify the actual effect of the new rules by building a virtual data environment. The simulation data environment should include various possible impact event scenarios, such as impacts of different intensities, impacts of different frequencies, and impact events at different locations. In this way, the adaptability and accuracy of the new rules can be comprehensively tested. The simulation test is conducted using statistical software. The software generates data based on preset impact parameters, and then processes the data using new rule parameters. Finally, evaluate the effectiveness of the new rules based on the processing results. If the new rules can effectively identify and respond to simulated impact events, then the new rules can be considered effective. At the same time, record the time and reason for modifying the rules to ensure that all relevant personnel understand the background and purpose of the rule adjustments. Through this series of detailed tests and records, the updated rule set is finally determined.

Claims

1. A condition monitoring system for road anti-collision guardrails, characterized in that: The system comprises: The data acquisition module collects sensor data of the anti-collision guardrail, reads the impact force, impact frequency and position information of each sensor, records the value and time stamp of the impact force, impact frequency and position information, verifies the data, excludes abnormal values, integrates all the impact force, impact frequency and position information data, generates a unified data format, and obtains real-time sensor data; A real-time transmission module receives the real-time sensor data, arranges the real-time sensor data in chronological order, caches the data stream, monitors the integrity and order of the data, detects whether there is data loss or duplication, marks the data, and generates a real-time data stream; An event recognition module analyzes the impact force, impact frequency and location information in the real-time data stream, compares the value of each parameter with a set threshold, determines whether the impact force exceeds the threshold, checks whether the impact frequency occurs multiple times in a short period of time, identifies the impact event that occurred, records the time, intensity and location of the event, and generates impact event information; A rule execution module reads the collision event information, checks the collision force, frequency and location information according to predefined rules, determines whether the parameters meet the rule conditions, generates an instruction to send an alarm if the collision force exceeds the target value and the frequency is higher than the set number of times, and generates a notification maintenance instruction if the location information is located in a critical section, executes sending an alarm, notifying maintenance and dispatching a maintenance team, and generates a response instruction set; A dynamic update module, based on the response instruction set and historical data, counts past impact events, analyzes the changing trends of impact force, frequency and position information, adjusts thresholds and conditions in predefined rules, updates rules in a rule base, records the modification time and reason of the rules, and obtains an updated rule set; The steps for obtaining the response instruction set are: Read the impact event information, analyze the impact force, impact frequency and position information, determine whether the impact force, impact frequency and position information meet predefined rule conditions, and obtain a rule matching result set; According to the data in the rule matching result set, determine which collision events meet the conditions of alarm or maintenance, such as the collision force exceeds the target value and the frequency is higher than the set number of times or the location is located in a critical section, and generate an early warning and maintenance decision set; Based on the warning and maintenance decision set, the formula is adopted: ； Calculate the emergency index of each decision, guide the decision of sending alarms, notifying maintenance or dispatching repair teams, and generate a response instruction set; Represents the degree of compliance of the alarm condition for each event, is the alarm weight, Represents the degree of compliance with maintenance conditions. To maintain weight; The steps of obtaining the update rule set are: Count past impact events, analyze the changing trends of impact force, impact frequency and location information, calculate the average change and standard deviation of parameters, and obtain a statistical analysis result set; Based on the statistical analysis result set, the need for rule adjustment is identified using the formula: ； Calculation rule parameters, where Indicates the current rule parameters, To adjust the intensity factor, and are the average changes in position and frequency, respectively, and are the standard deviations of position and frequency, respectively, is the standard deviation adjustment coefficient and generates a new rule candidate set.

2. The state monitoring system for road anti-collision guardrail according to claim 1, characterized in that: The steps for acquiring the real-time sensor data are as follows: Collect data from each sensor on the anti-collision guardrail, including impact force, impact frequency and location information, record the data value and timestamp, integrate and generate impact data in a unified format to obtain an impact data set; Verifying and eliminating outliers on the collision data set, using a cleaning algorithm to process abnormal data points, ensuring data quality and integrity, and generating a cleaned collision data set; The cleaned impact data set is integrated to aggregate all impact force, impact frequency and position information data using the formula: ； Calculate the weighted impact force index to obtain real-time sensor data, where: Represents the impact force, is the impact force weight, represents the impact frequency, is the frequency weight, Represents location information, is the position weight.

3. The state monitoring system for road crash barriers according to claim 1, characterized in that: The steps of acquiring the real-time data stream are: Receiving the real-time sensor data, arranging the data in chronological order, maintaining the continuity of the timeline and the timing correctness of the data, and generating a time-ordered data set; caching the time-ordered data set, monitoring the integrity and order of the data stream in real time, checking whether the data is lost or duplicated, marking abnormal data points, and generating a monitored data stream; From the monitored data stream, all repeated or lost data events are detected and marked, and the repeated data are cleared to obtain an optimized real-time data stream.

4. The state monitoring system for road crash barriers according to claim 3, characterized in that: The steps to identify a collision event are: Analyzing the optimized real-time data stream, comparing the impact force, impact frequency and position information of each data point with a preset threshold, marking the data points exceeding the threshold, and obtaining an over-threshold data set; From the above-threshold data set, the impact events with a frequency exceeding the threshold in a short period of time are identified using the formula: ； Calculate which events meet the dual conditions of impact force and frequency, and generate a preliminary impact event identification set, where Represents the impact force, is the impact force threshold, represents the impact frequency, is the frequency threshold; The preliminary impact event identification set is refined, the location information is analyzed, and it is verified which impact event locations are key areas. The time, intensity and location information are integrated to complete the identification of the impact event.

5. The state monitoring system for road crash barriers according to claim 1, characterized in that: The steps to determine whether the impact force exceeds the threshold are: Analyzing the data in the real-time data stream, including the impact force, impact frequency and position information, and comparing them with preset thresholds to determine whether each parameter exceeds the threshold, and generating a threshold comparison data set; The threshold comparison data set is analyzed using the formula: ； Calculate the deviation of each data point from the threshold, where Represents the impact force, is the impact force threshold, represents the impact frequency, For the frequency threshold, a deviation calculation result set is generated; From the deviation calculation result set, the data point with the largest deviation is selected to determine the data point with the largest deviation as a high-risk candidate for the impact event, and the impact force, frequency and position information of the data point with the largest deviation is recorded and sorted to obtain the impact event threshold comparison result.

6. The state monitoring system for road crash barriers according to claim 1, characterized in that: The step of acquiring the updated rule set also includes: evaluating the effect of the new rule candidate set, verifying the accuracy and effect of the new rule through simulation data, recording the time and reason of rule modification, and obtaining the updated rule set.

Images