A distributed environment parameter intelligent early warning method based on total control network stream computing

Abstract

The application discloses a kind of distributed environmental parameter intelligent early warning methods based on total control net stream computing, it is related to activity launch platform environmental intelligent early warning technical field, this method is by dividing into multiple monitoring points to launch platform environment, and the running state data of environment and key equipment are collected;Calculate the priority coefficient of monitoring point, judge whether key safety parameter meets the requirement;Under the framework of stream computing, establish minute level and hour level sliding window, extract multi-time scale parameter fluctuation characteristics, calculate stability index and compare with threshold value analysis;For adjacent monitoring node, calculate multidimensional parameter consistency and obtain overall consistency index, realize single point link fault and global environmental anomaly discrimination;Automatic link traceability log is generated, which provides basis and parameter optimization correction for abnormal event tracing.The method can realize efficient monitoring and intelligent early warning of multi-point environment and equipment parameters of activity launch platform, improve the accuracy and timeliness of safety protection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent early warning of mobile launch platform environment, in particular to a distributed environment parameter intelligent early warning method based on total control network stream computing. BACKGROUND

[0002] As the core equipment for performing vertical transportation, launch support and attitude adjustment of the launch vehicle, the operation safety of the mobile launch platform is directly related to the reliable execution of the launch task. The platform integrates multiple complex systems such as machinery, electrical control, hydraulic pressure, walking and gas guiding and discharging, and the internal operating environment is affected by multiple factors such as external meteorological conditions, equipment heating, hydraulic pressure fluctuation and air flow disturbance. During the operation of the platform, there are rapid fluctuations and coupled changes of multi-dimensional parameters such as temperature, humidity, wind speed, hydraulic pressure and electrical load. If the monitoring and early warning are not timely, the platform support structure may be unstable, the hydraulic system may be abnormal, the electronic equipment may malfunction or the attitude may deviate, thereby threatening the launch safety.

[0003] The existing launch platform mostly uses periodic collection and manual inspection, which has the following disadvantages:

[0004] Response lag, manual inspection and timed sampling method cannot capture the second-level environmental and system fluctuations, and cannot meet the high dynamic monitoring requirements in the launch preparation stage;

[0005] Insufficient monitoring coverage, limited sensor layout points, and unable to realize the global state perception of different areas of the platform (such as umbilical tower, hydraulic support area, guiding and discharging system and walking mechanism);

[0006] Data processing lag, traditional system uses batch computing method, which is difficult to perform stream analysis and real-time abnormality identification on multi-source asynchronous data;

[0007] Insufficient intelligent decision-making, lack of correlation analysis and multi-node consistency judgment mechanism among temperature, humidity, wind speed, electrical load and hydraulic parameters, resulting in high false alarm rate and slow strategy response. SUMMARY

[0008] In view of the deficiencies of the prior art, the present application provides a distributed environment parameter intelligent early warning method based on total control network stream computing to solve the problems mentioned in the background art.

[0009] To achieve the above purpose, the present application realizes the following technical scheme: a distributed environment parameter intelligent early warning method based on total control network stream computing, comprising the following steps:

[0010] Step one, by dividing the active launch platform environment into multiple monitoring points, real-time collection of the i th monitoring point running data, including: environmental temperature data, environmental humidity data, wind speed data and equipment running state data; preprocessing the collected data, establishing the parameter set of the i th monitoring point, and inputting the platform total control network in an event-driven manner;

[0011] Step two, by extracting the normalized value of the parameter set of the i th monitoring point, calculating the priority coefficient Pyx, and comparing and analyzing with the priority threshold Pth, judging whether the parameter flow of the current monitoring point meets the key safety parameter requirement, if not, giving the corresponding strategy;

[0012] Step three, by establishing minute and hour level sliding window under the streaming computing framework of the i th monitoring point, segmenting and real-time updating the environmental and equipment parameters, capturing the fluctuation characteristics of different time scales, obtaining parameter sequence values XTm and XTh; based on sliding window parameter calculation stability index SW, and comparing and analyzing with stability threshold Sth, judging whether the matching degree of the current parameter sequence XTm and XTh is qualified, if not, giving the corresponding strategy;

[0013] Step four, by vectorizing, normalizing and calculating the consistency of multi-dimensional environmental parameters of adjacent monitoring nodes, then synthesizing each dimension to obtain the overall consistency index Cij of the i th monitoring point and the j th monitoring point, and comparing and analyzing with the arbitration threshold Cth, judging whether the data of the i th monitoring node and the j th monitoring node is consistent, if not, it is identified as single point link failure, and the global alarm is not triggered; if it is consistent, it is identified as global environmental abnormality, and the corresponding strategy is given;

[0014] Step five, through the whole process of early warning instruction automatic generation of link traceability log, complete recording of parameter collection, priority determination, window comparison and consistency arbitration steps, for accident tracing and parameter optimization.

[0015] Preferably, the step one comprises:

[0016] S11, divide the entire active launch platform environment into several monitoring points, n monitoring points in the platform environment monitoring module are sequentially marked according to the number Z1, Z2, Z3,..., Zn; n Real-time monitoring is carried out on the running process of the i th monitoring point; the running data of the i th monitoring point includes: environmental temperature data, environmental humidity data, wind speed data and equipment running state data;

[0017] S111, real-time monitoring of the umbilical tower base structure area and the hydraulic support area operation process of the i th monitoring point of the mobile launch platform; installing high-temperature-resistant and vibration-resistant environmental sensors near the umbilical tower base, the hydraulic cylinder shell and the support arm; real-time collection and recording of environmental temperature parameters;

[0018] S112, real-time monitoring of the electrical control cabin and the enclosed channel operation process of the i th monitoring point of the mobile launch platform; arranging industrial-grade humidity sensors in the air outlet of the electrical cabin, the cable trench and the sealed room area; real-time collection and recording of environmental humidity parameters;

[0019] S113, real-time monitoring of the gas guide and exhaust port, the platform contour and the air inlet of the walking system operation process of the i th monitoring point of the mobile launch platform; installing three-dimensional ultrasonic wind speed and direction instruments in the guide and exhaust port, the platform contour air hole and the walking chassis air inlet; real-time collection of wind speed parameters;

[0020] S114, real-time monitoring of the hydraulic system and power distribution area operation process of the i th monitoring point of the mobile launch platform; installing current, voltage and vibration integrated sensors at the hydraulic pump station, the accumulator outlet, the switch cabinet and the key cable outlet; real-time collection of equipment operation state parameters, including current, voltage, hydraulic pressure and abnormal vibration signals.

[0021] Preferably, the step one also includes:

[0022] S12, data preprocessing of the environmental temperature parameters, the environmental humidity parameters, the wind speed parameters and the electrical equipment operation state parameters;

[0023] S121, using Kalman filtering and median filtering algorithm to suppress noise and remove abnormal points of temperature parameters and humidity parameters; removing high-frequency interference components of wind speed parameters through fast Fourier transform FFT analysis; using wavelet denoising method to eliminate pulse interference in current, voltage and hydraulic sampling of equipment operation state parameters;

[0024] S122, using Z-score standardization method to normalize and unify the dimension of the environmental temperature parameters, the environmental humidity parameters, the wind speed parameters and the equipment operation state parameters; using NTP network time synchronization protocol to mark each sampling data with millisecond time stamp, so that the time sequence consistency of multi-source data under the stream computing framework is achieved;

[0025] S13, establishing the parameter set of the i th monitoring point by MQTT protocol and Kafka stream message queue, and inputting the platform general control network in an event-driven manner.

[0026] Preferably, the step two includes:

[0027] S21, extract the environment temperature parameter normalized value Tg, the environment humidity parameter normalized value Hd, the wind speed parameter normalized value Vs and the equipment running state parameter normalized value Ey in the parameter set of the i-th monitoring point, and calculate the priority coefficient Pyx.

[0028] Preferably, the step two further comprises:

[0029] S22, by presetting the priority threshold Pth, and comparing and analyzing the priority coefficient Pyx with the priority threshold Pth, the first evaluation result is obtained, which comprises:

[0030] When the priority coefficient Pyx is greater than the priority threshold Pth, it indicates that the parameter flow of the current monitoring point meets the key safety parameter requirement, and automatically enters the high priority channel, so that the current parameter flow is transmitted and processed within a second-level time delay;

[0031] When the priority coefficient Pyx is less than or equal to the priority threshold Pth, it indicates that the parameter flow of the current monitoring point does not meet the key safety parameter requirement, a first warning instruction is triggered, and a first strategy is generated: the current parameter flow is placed in the ordinary channel for transmission in a batch manner or a longer transmission cycle; a monitoring enhancement mechanism is started to increase the sampling frequency and data cache depth of the current monitoring point to prevent abnormal fluctuations from being ignored; if the parameter flow does not meet the key safety parameter requirement for three consecutive times, an artificial inspection prompt is triggered to prompt the operation and maintenance personnel to check the sensor state and data link.

[0032] Preferably, the step three comprises:

[0033] S31, under the stream computing framework of the i-th monitoring point, a minute-level sliding window Wm and an hour-level sliding window Wh are respectively established, the environment temperature, the environment humidity, the wind speed and the equipment running state parameters are stored and updated in segments in real time, and the parameter fluctuation characteristics under different time scales are captured to obtain the parameter sequence values XTm and XTh of the minute-level sliding window Wm and the hour-level sliding window Wh.

[0034] Preferably, the step three further comprises:

[0035] S32, the parameter sequence values XTm and XTh in the minute-level sliding window Wm and the hour-level sliding window Wh are extracted, the average values are respectively calculated, and after dimensionless processing, the stability index SW is calculated.

[0036] S33, by presetting the stability threshold Sth, and comparing and analyzing the stability index SW with the stability threshold Sth, the second evaluation result is obtained, which comprises:

[0037] When the stability index SW is less than or equal to the stability threshold Sth, it indicates that the matching degree of the current parameter sequence values XTm and XTh is qualified, and the monitoring is continued.

[0038] When the stability index SW is greater than the stability threshold Sth, it indicates that the matching degree of the current parameter sequence value XTm and XTh is not qualified, there is a risk of short-term sudden change and trend deviation; the second early warning instruction is triggered, and the second strategy is generated: the current monitoring point is marked as a key data source, and the scheduling priority in the total control network is improved; at the same time, cross verification is performed on the short window and the long window data, and incidental noise and single-point sensor distortion are eliminated; the sampling frequency and the data points are encrypted in the next two monitoring periods; if the stability index SW is greater than the stability threshold Sth for three consecutive monitoring periods, the consistency arbitration detection mechanism is started.

[0039] Preferably, the step four comprises:

[0040] S41, when the stability index SW is greater than the stability threshold Sth for three consecutive monitoring periods, the consistency arbitration detection mechanism is started, the multi-dimensional environmental parameter data of the ith monitoring node and the adjacent jth monitoring node in the same time slice is extracted, the vectorization representation method is used, and the parameter dimension is constructed into a multi-parameter vector form to obtain a node parameter vector set X i and X j .

[0041] Preferably, the step four further comprises:

[0042] S42, by using the node parameter vector set X i and X j , the parameter vector is standardized by using a normalization algorithm, the differences in different parameter dimensions and orders of magnitude are eliminated, the relative difference degree calculation method is used to perform consistency calculation on the node parameter pair in each dimension, and the overall consistency index C ij of the ith monitoring point and the jth monitoring point is obtained by using a weighted average algorithm to synthesize the results of each dimension.

[0043] S43, by presetting an arbitration threshold Cth, and comparing and analyzing the overall consistency index C ij with the arbitration threshold Cth, a third evaluation result is obtained, including:

[0044] When the overall consistency index C ij is less than the arbitration threshold Cth, it indicates that the data of the ith monitoring node and the jth monitoring node is inconsistent, and it is determined as a single-point link fault, without triggering a global alarm; manual inspection or a backup sensing node is notified;

[0045] When the overall consistency index C ijWhen greater than arbitration threshold Cth, it indicates that the i-th monitoring node and the j-th monitoring node are consistent with data, and is determined as a global environmental anomaly, a third early warning instruction is triggered, and a third strategy is generated: the current parameter flow is included in a high-priority channel to ensure second-level transmission and processing; the early warning is pushed to a platform general control network; rapid protection measures including local load reduction, auxiliary cooling and exhaust device activation and risk area isolation are executed at the edge node; the current abnormal event and consistency index determination result are recorded to a log database.

[0046] Preferably, the step five comprises:

[0047] S51, through the whole process triggered by the first, second and third early warning instructions, a complete link tracing log is automatically generated, and each step from parameter collection, priority determination, window comparison and consistency arbitration is recorded in detail; the log is used for providing basis for accident tracing and for parameter and model optimization correction.

[0048] The application provides a distributed environmental parameter intelligent early warning method based on general control network stream computing.

[0049] (1) The distributed environmental parameter intelligent early warning method based on general control network stream computing realizes millisecond-level collection and dynamic update of environmental parameters by deploying a distributed stream computing framework at multiple monitoring points and combining minute-level and hour-level sliding window mechanisms, so that the platform can quickly identify temperature, humidity and wind speed mutations during fuel filling, equipment debugging and pre-launch high-load stages, and the real-time performance and response speed of environmental monitoring are significantly improved.

[0050] (2) The distributed environmental parameter intelligent early warning method based on general control network stream computing realizes a dynamic scheduling mechanism that automatically enters high-risk parameters into a high-priority channel by establishing a parameter priority coefficient model, calculating the importance of parameters in real time according to the weight relationship of temperature, humidity, wind speed and equipment state, and comparing with the priority threshold, so that critical safety parameters can be transmitted and processed within a second-level delay during the launch preparation stage, and delayed early warning is effectively avoided.

[0051] (3) The distributed environmental parameter intelligent early warning method based on general control network stream computing realizes accurate diagnosis of the monitoring link by vectorization and normalization analysis of multi-dimensional environmental parameters of adjacent monitoring nodes through a consistency arbitration detection mechanism, calculates the overall consistency index and automatically distinguishes between "local fault" and "global anomaly", so that the whole process of abnormal events can be traced back, and quantitative basis is provided for subsequent accident analysis and model optimization.

[0052] (4) The distributed environmental parameter intelligent early warning method based on the total control network streaming computing can automatically trigger edge node rapid protection measures when the system is determined to be a global environmental anomaly, including local load reduction, activation of auxiliary cooling and ventilation devices and risk area isolation, so as to realize the linkage closed-loop control of environmental parameter anomaly and equipment protection, thereby effectively improving the platform's safety redundancy and system resilience under extreme climate or abnormal vibration. Attached Figure Description

[0053] Figure 1 This is a schematic diagram illustrating the steps of a distributed intelligent early warning method for environmental parameters based on streaming computation of a central control network, according to the present invention. Detailed Implementation

[0054] 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.

[0055] Example 1: Please refer to Figure 1 This invention provides a distributed intelligent early warning method for environmental parameters based on streaming computing of a central control network, comprising the following steps:

[0056] Step 1: Divide the environment of the active launch platform into multiple monitoring points, and collect the operating data of the i-th monitoring point in real time, including: ambient temperature data, ambient humidity data, wind speed data, and equipment operating status data; preprocess the collected data, establish the parameter set of the i-th monitoring point, and input it into the platform's overall control network in an event-driven manner;

[0057] Step 2: By extracting the normalized values ​​of the parameters in the parameter set of the i-th monitoring point, calculate the priority coefficient Pyx and compare it with the priority threshold Pth to determine whether the parameter flow of the current monitoring point meets the key safety parameter requirements. If it does not meet the requirements, give the corresponding strategy.

[0058] Step 3: Under the streaming computing framework of the i-th monitoring point, establish minute-level and hour-level sliding windows to segment and update environmental and equipment parameters in real time, capture fluctuation characteristics at different time scales, and obtain parameter sequence values ​​XTm and XTh; calculate the stability index SW based on the sliding window parameters, and compare it with the stability threshold Sth to determine whether the matching degree between the current parameter sequence XTm and XTh is qualified. If it is not qualified, an appropriate strategy is given.

[0059] Step 4: By vectorizing and normalizing the multidimensional environmental parameters of adjacent monitoring nodes and calculating the dimension-by-dimensional consistency, the overall consistency index Cij between the i-th and j-th monitoring points is obtained by integrating all dimensions. This index is then compared with the arbitration threshold Cth to determine whether the data of the i-th and j-th monitoring nodes are consistent. If they are inconsistent, it is considered a single-point link failure and no global alarm is triggered. If they are consistent, it is considered a global environmental anomaly and an appropriate strategy is applied.

[0060] Step 5: Automatically generate a link tracing log throughout the entire process using early warning commands, fully recording the steps of parameter collection, priority determination, window comparison, and consistency arbitration, for use in accident tracing and parameter optimization.

[0061] In this embodiment, by introducing multi-level sliding window analysis and node consistency arbitration mechanism into the environmental monitoring process of the active launch platform, the temporal characteristics of monitoring data are captured, spatial consistency is determined, and the entire process is traceable and managed. This can effectively improve the environmental monitoring accuracy and anomaly identification reliability of the launch platform under complex climate and high dynamic conditions, thereby ensuring the rapid response and traceability of the system in the process of early warning, diagnosis and safety decision-making.

[0062] Example 2: This example is an explanation of Example 1. Please refer to the example provided. Figure 1 Specifically, step one includes:

[0063] S11. Divide the entire launch platform environment into several monitoring points. These n monitoring points are numbered Z1, Z2, Z3, ..., Z... in the platform environment monitoring module. n Sequential marking is performed; the operation process of the i-th monitoring point is monitored in real time; the operation data of the i-th monitoring point includes: ambient temperature data, ambient humidity data, wind speed data, and equipment operation status data;

[0064] S111. Real-time monitoring of the operation of the umbilical tower base structure area and hydraulic support area at the i-th monitoring point of the active launch platform; installation of high-temperature and vibration-resistant environmental sensors near the umbilical tower base, hydraulic cylinder shell and support arm; real-time acquisition and recording of environmental temperature parameters;

[0065] S112. Real-time monitoring of the operation of the electrical control cabin and enclosed passage at the i-th monitoring point of the active launch platform; deployment of industrial-grade humidity sensors in the air outlet, cable trench and sealed area of ​​the electrical cabin; real-time collection and recording of environmental humidity parameters;

[0066] S113, Real-time monitoring of the running process of the gas guide port, platform contour and walking system air inlet of the i th monitoring point of the mobile launching platform; installing a three-dimensional ultrasonic wind speed and direction instrument on the guide port, platform contour air hole and walking chassis ventilation port; real-time collection of wind speed parameters;

[0067] S114, Real-time monitoring of the running process of the i th monitoring point of the mobile launching platform hydraulic system and power distribution area; installing current, voltage and vibration integrated sensors on the hydraulic pump station, accumulator outlet, switch cabinet and key cable outlet; real-time collection of equipment running state parameters, including current, voltage, hydraulic pressure and abnormal vibration signals.

[0068] In this embodiment, by arranging multiple types of high-precision sensors in key areas such as the umbilical tower seat, electrical control cabin, gas guide port and hydraulic system of the mobile launching platform, the global synchronous collection and partition monitoring of multi-source parameters such as temperature, humidity, wind speed and equipment state are realized, which can effectively reflect the changes of the environment and running state of different structural areas of the platform, significantly improve the spatial resolution and data integrity of environmental monitoring, and provide more accurate and reliable basic data support for subsequent abnormal identification and safety warning.

[0069] Embodiment 3: This embodiment is an explanation and description in embodiment 2, please refer to Figure 1 , specifically, the step one further comprises:

[0070] S12, data preprocessing of the environmental temperature parameter, environmental humidity parameter, wind speed parameter and electrical equipment running state parameter;

[0071] S121, using Kalman filter and median filter algorithm to suppress noise and remove abnormal points of temperature parameter and humidity parameter; removing high-frequency interference components of wind speed parameter by fast Fourier transform FFT analysis; using wavelet denoising method to eliminate pulse interference in current, voltage and hydraulic sampling of device running state parameter;

[0072] S122, using Z-score standardization method to normalize the environmental temperature parameter, environmental humidity parameter, wind speed parameter and equipment running state parameter, and unify the dimension; using NTP network time synchronization protocol to mark each sampling data with millisecond time stamp, so that the time sequence consistency of multi-source data under the stream computing framework is realized;

[0073] S13, through the MQTT protocol and Kafka stream message queue, the processed environmental temperature parameter, environmental humidity parameter, wind speed parameter and equipment running state parameter are established as the parameter set of the i th monitoring point, and are input into the platform general control network in an event-driven manner.

[0074] In this embodiment, by introducing the preprocessing mechanism of Kalman filtering, FFT frequency domain denoising, wavelet denoising and Z-score normalization in the active launch platform environment monitoring process, and combining MQTT and Kafka streaming message queue to realize the unified access and millisecond-level synchronization of multi-source heterogeneous data, the stability and timing consistency of data transmission can be maintained in strong electromagnetic interference, high vibration and complex airflow environment, which significantly improves the real-time monitoring accuracy and anti-interference ability of the active launch platform in high temperature, high humidity, strong wind and transient electromagnetic impact working conditions, and ensures the reliability and continuity of the whole cycle environment monitoring of the launch task.

[0075] Embodiment 4: This embodiment is an explanation and description made in Embodiment 3, please refer to Figure 1 , specifically, the step two comprises:

[0076] S21, extracting the environmental temperature parameter normalized value Tg, the environmental humidity parameter normalized value Hd, the wind speed parameter normalized value Vs and the device running state parameter normalized value Ey in the parameter set of the ith monitoring point, calculating and obtaining the priority coefficient Pyx, the formula is as follows:

[0077] ;

[0078] In the formula, w1, w2, w3 and w4 represent weight coefficients.

[0079] w1=0.35: representing the influence of temperature parameter on the priority of monitoring point parameters of active launch platform, occupying a higher weight, being a key indicator, directly reflecting the sensitivity of temperature change of launch operation environment to hydraulic system, electrical cabin and structural safety;

[0080] w2=0.25: representing the influence of humidity parameter on the priority of monitoring point parameters, occupying a medium weight, reflecting the regulation and control contribution of humidity to the insulation performance and corrosion risk of electrical equipment, which plays an important role in the safe operation of the electrical control system of the launch platform;

[0081] w3=0.3: representing the influence of wind speed parameter on the priority of monitoring point parameters, occupying a secondary high weight, reflecting the auxiliary role of ventilation and airflow conditions on gas guiding and discharging, heat dissipation and external contour stability;

[0082] w4=0.1: representing the influence of device running state parameter on the priority of monitoring point parameters, occupying an auxiliary weight, reflecting the auxiliary contribution of current, voltage and vibration signals to the running stability of the overall power system and support structure;

[0083] By constructing the priority coefficient Pyx weighted by w1, w2, w3 and w4, the risk level of the monitoring point under different environments and device states can be quantified, providing a scientific basis for streaming calculation scheduling and safety warning.

[0084] In this embodiment, by normalizing the environmental temperature, humidity, wind speed and equipment running state parameters collected by each monitoring point of the active launch platform, and constructing a priority coefficient Pyx with a weight coefficient, the quantitative evaluation and dynamic sorting of multi-dimensional environment and equipment state are realized. The risk level of the key monitoring point can be automatically identified according to the task stage. This method can quickly focus on the high-risk parameter change area in different stages such as launch preparation, gas guiding and discharging, ignition and recovery, realize intelligent allocation and real-time control of monitoring resources, and significantly improve the safety warning response speed and decision accuracy of the active launch platform in complex environment.

[0085] Embodiment 5: This embodiment is an explanation and description in embodiment 4, please refer to Figure 1 , specifically, the step two further comprises:

[0086] S22, by comparing and analyzing the priority coefficient Pyx and the priority threshold Pth through the preset priority threshold Pth, a first evaluation result is obtained, including:

[0087] When the priority coefficient Pyx is greater than the priority threshold Pth, it indicates that the parameter flow of the current monitoring point meets the key safety parameter requirement, and automatically enters the high priority channel, so that the current parameter flow is transmitted and processed within a second-level time delay;

[0088] When the priority coefficient Pyx is less than or equal to the priority threshold Pth, it indicates that the parameter flow of the current monitoring point does not meet the key safety parameter requirement, a first warning instruction is triggered, and a first strategy is generated: the current parameter flow is placed in the ordinary channel for transmission in batch mode or long transmission cycle; start the monitoring enhancement mechanism to increase the sampling frequency and data cache depth of the current monitoring point to prevent abnormal fluctuations from being ignored; if the parameter flow does not meet the key safety parameter requirement for three consecutive times, an artificial inspection prompt is triggered to prompt the operation and maintenance personnel to check the sensor state and data link.

[0089] The acquisition method of the priority threshold Pth: by statistically analyzing the historical data of each monitoring point in the running process of the active launch platform, the parameter priority coefficient distribution range under various environmental conditions (high wind, high temperature, high humidity, filling working condition) and different equipment running states is extracted, and combined with the experience judgment of platform operation and maintenance and launch engineering experts, a reasonable priority threshold is determined; this threshold is used to distinguish whether the parameter flow should enter the high priority channel in the transmission and processing process, so as to ensure the transmission immediacy and warning accuracy of the key safety parameters in the high dynamic launch preparation stage.

[0090] In this embodiment, by setting the priority threshold Pth and dynamically comparing the priority coefficient Pyx, the hierarchical transmission and intelligent scheduling of the monitoring data of the mobile launch platform are realized. When the monitoring point is in a high-risk working condition such as high temperature, strong wind or fuel filling, the key safety parameters can automatically enter the high-priority channel, and the data transmission and processing are completed within a second-level time delay; while the data in the ordinary state is transmitted in batches, reducing the system bandwidth occupation. This mechanism significantly improves the response speed and resource utilization efficiency of the environment and equipment monitoring in the launch preparation stage, ensures the real-time and reliability of the key safety information, and avoids the safety hazards caused by data delay.

[0091] Embodiment 6: This embodiment is an explanation in embodiment 5, please refer to Figure 1 , specifically, the step three comprises:

[0092] S31, under the streaming computing framework of the i-th monitoring point, a minute-level sliding window Wm and an hour-level sliding window Wh are respectively established, the environmental temperature, the environmental humidity, the wind speed and the equipment running state parameters are stored and updated in segments, and the parameter fluctuation characteristics under different time scales are captured, and the parameter sequence values XTm and XTh of the minute-level sliding window Wm and the hour-level sliding window Wh are obtained.

[0093] In this embodiment, by establishing the minute-level sliding window Wm and the hour-level sliding window Wh under the streaming computing framework of the i-th monitoring point of the mobile launch platform, the dynamic change characteristics of the temperature, humidity, wind speed and equipment state can be captured at different time scales. This design can identify long-period trend changes such as equipment cabin temperature rise or humidity accumulation while monitoring sudden environmental fluctuations such as heat flow impact or airflow turbulence in the countdown phase of the launch. Thus, multi-scale evaluation of the environmental stability of the key area of the platform is realized, the abnormal identification accuracy and response timeliness in the launch preparation and countdown phase are improved, and real-time, continuous and reliable data support is provided for safe launch.

[0094] Embodiment 7: This embodiment is an explanation in embodiment 6, please refer to Figure 1 , specifically, the step three further comprises:

[0095] S32, the parameter sequence values XTm and XTh in the minute-level sliding window Wm and the hour-level sliding window Wh are extracted, the average values are respectively calculated, and after dimensionless processing, the stability index SW is calculated, and the formula is as follows:

[0096] ;

[0097] S33, by presetting the stability threshold Sth and comparing the stability index SW with the stability threshold Sth, a second evaluation result is obtained, including:

[0098] When the stability index SW is less than or equal to the stability threshold Sth, it indicates that the matching degree of the current parameter sequence value XTm and XTh is qualified, and continuous monitoring is performed.

[0099] When the stability index SW is greater than the stability threshold Sth, it indicates that the matching degree of the current parameter sequence value XTm and XTh is unqualified, there is a risk of short-term sudden change and trend deviation, a second early warning instruction is triggered, and a second strategy is generated: the current monitoring point is marked as a key data source, and the scheduling priority of the key data source in the total control network is improved; at the same time, cross verification is performed on the short window and the long window data to eliminate accidental noise and single-point sensor distortion; the sampling frequency and the data points are increased in the next two monitoring periods; if the stability index SW is greater than the stability threshold Sth for three consecutive monitoring periods, a consistency arbitration detection mechanism is started.

[0100] The stability threshold Sth is obtained by: statistically analyzing the historical monitoring data in the running process of the active launch platform, extracting the stability index SW distribution range under normal running state and abnormal working condition, and combining the experience judgment of the launch platform running expert to determine a reasonable stability threshold; referring to the launch safety specification and historical abnormal event data, the threshold is used to distinguish the stability state and sudden abnormality of the parameter.

[0101] In this embodiment, by introducing the calculation and threshold determination mechanism of the stability index SW in the monitoring process of the active launch platform, the joint identification of short-term fluctuation and long-term trend deviation of environmental and equipment parameters can be realized. When the parameter sequence in the minute and hour window deviates significantly, the system can quickly trigger the second early warning instruction and perform dynamic scheduling and sampling encryption, so as to discover potential risks such as temperature sudden rise, humidity abnormality or air flow disturbance in advance during the launch preparation and fuel filling stage. The mechanism effectively improves the parameter stability monitoring capability of the launch platform in a high dynamic and high coupling environment, realizes early identification and intervention of abnormal fluctuation, and provides real-time decision basis for launch safety guarantee. Figure 1 , specifically, the step four comprises:

[0102] S41, when the stability index SW is greater than the stability threshold Sth for three consecutive monitoring periods, a consistency arbitration detection mechanism is started, multi-dimensional environmental parameter data of the ith monitoring node and the adjacent jth monitoring node in the same time slice is extracted, a vectorization representation method is used, and a node parameter vector set X i and X j is obtained by constructing a multi-parameter vector form according to the parameter dimension.

[0103] In this embodiment, by introducing inter-node consistency arbitration detection mechanism in the active launch platform, when stability anomaly appears in continuous multiple monitoring periods, the system automatically extracts multi-dimensional environmental parameters of adjacent monitoring nodes and performs vector analysis, which can realize synchronous comparison and judgment across regions and systems. This mechanism can distinguish between local sensor failure and overall environmental disturbance, improving the accuracy and robustness of anomaly identification; it is especially suitable for complex scenarios with strong coupling effects between the launch frame, fuel supply area and control cabin, and helps to identify global environmental risks in key platform operation stages, ensuring the stability and safety of the launch system.

[0104] Embodiment 9: This embodiment is an explanation and description in embodiment 8, please refer to Figure 1 , specifically, the step four further comprises:

[0105] S42, through the node parameter vector set X i and X j , using a normalization algorithm to standardize the parameter vector dimension by dimension, eliminating the difference between different parameter dimensions and orders of magnitude, using a relative difference calculation method, calculating the consistency of the node parameter pair dimension by dimension, and through a weighted average algorithm, the overall consistency index C ij of the i th monitoring point and the j th monitoring point is obtained.

[0106] ;

[0107] In the formula, k represents the total number of parameter dimensions, represents the single parameter consistency value of the i th and j th monitoring nodes on the p th parameter dimension;

[0108] ;

[0109] In the formula, represents the value of the i th monitoring node in the node parameter vector set X i on the p th parameter dimension, represents the value of the j th monitoring node in the node parameter vector set X j on the p th parameter dimension, represents a small constant to prevent the denominator from being zero;

[0110] ;

[0111] ;

[0112] S43, by presetting the arbitration threshold Cth, and comparing the overall consistency index C ij with the arbitration threshold Cth, the third evaluation result is obtained, including:

[0113] When the overall consistency index C ij When the arbitration threshold Cth, it indicates that the i-th monitoring node and the j-th monitoring node data are inconsistent, it is determined as a single-point link failure, and no global alarm is triggered; manual inspection or standby sensing node is notified;

[0114] When the overall consistency index C ij When the arbitration threshold Cth, it indicates that the i-th monitoring node and the j-th monitoring node data are consistent, it is determined as a global environmental anomaly, a third early warning instruction is triggered, and a third strategy is generated: the current parameter flow is included in a high-priority channel to ensure second-level transmission and processing; the warning is pushed to the platform control network; rapid protection measures including local load reduction, auxiliary cooling and exhaust device and isolation of risk area are executed at the edge node; the current abnormal event and consistency index determination result are recorded to the log database.

[0115] The arbitration threshold Cth is obtained by: statistically analyzing a large amount of historical running data of the monitoring nodes, extracting the distribution interval of the consistency index between the nodes under different working conditions, combining expert experience to set a reasonable arbitration threshold, and referring to the launch platform safety specification and historical accident case data to correct the threshold to ensure that the determination can effectively distinguish between sensor local failure and real working condition anomaly, and ensure the timeliness and accuracy of the warning.

[0116] In this embodiment, by using multi-dimensional parameter normalization and overall consistency index calculation between nodes in the active launch platform, accurate consistency evaluation of adjacent monitoring points can be realized. When the consistency index reaches the preset arbitration threshold, the system can quickly identify global environmental anomalies and automatically trigger high-priority channel transmission and edge node protection measures; otherwise, when the consistency index is lower than the threshold, single-point link failure can be distinguished to avoid false triggering of global alarm. This mechanism in the platform fuel area, control cabin and hydraulic system and other highly coupled complex scenes helps to ensure the rapid response and safety protection of abnormal events and improve the reliability and safety of the overall operation of the launch platform.

[0117] Embodiment 10: This embodiment is an explanation and description in embodiment 9, please refer to Figure 1 , specifically, the step five comprises:

[0118] S51, through the whole process triggered by the first, second and third early warning instructions, a complete link tracing log is automatically generated, which details the steps from parameter collection, priority determination, window comparison and consistency arbitration; the log is used as a basis for accident tracing and for parameter and model optimization correction.

[0119] In this embodiment, by automatically generating the whole-process link traceability log in the active launch platform, the processing flow of parameter acquisition, priority determination, sliding window comparison and consistency arbitration of each monitoring point can be recorded completely, accurate basis for abnormal event tracing is provided, subsequent parameter optimization and model correction are supported, and the platform safety management, fault analysis and operation and maintenance efficiency are significantly improved in the launch preparation and operation process.

[0120] The size of the threshold is set for comparison, and the size of the threshold depends on how much sample data and the number of base set by the person skilled in the art for each group of sample data; as long as it does not affect the proportional relationship between the parameters and the quantized values.

[0121] The above formulas are obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value, and the coefficients in the formula are set by the person skilled in the art according to the actual situation. The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art within the technical range disclosed by the present application, according to the technical scheme and the inventive concept of the present application, should be covered within the protection scope of the present application.

Claims

1. A method for intelligent early warning of environmental parameters based on total control network stream computing, characterized in that, The method comprises the following steps: Step one, by dividing the active launch platform environment into multiple monitoring points, real-time collection of the i th monitoring point operation data, including: environmental temperature data, environmental humidity data, wind speed data and equipment operation state data; preprocessing the collected data, establishing the parameter set of the i th monitoring point, and inputting the platform total control network in an event-driven manner; Step two, by extracting the normalized value of the parameters in the parameter set of the i th monitoring point, calculating the priority coefficient Pyx, and comparing and analyzing it with the priority threshold Pth, it is judged whether the parameter flow of the current monitoring point meets the key safety parameter requirement, and if not, the first strategy is given; Step three, by establishing a minute-level and hour-level sliding window under the streaming computing framework of the i th monitoring point, the environmental and equipment parameters are stored and updated in segments in real time, the fluctuation characteristics of different time scales are captured, and the parameter sequence values XTm and XTh are obtained; based on the sliding window parameter calculation, the stability index SW is calculated, and compared with the stability threshold Sth, it is judged whether the matching degree of the current parameter sequence XTm and XTh is qualified, and if not, the second strategy is given; Step four, by vectorizing, normalizing and calculating the consistency of each dimension of the multi-dimensional environmental parameters of adjacent monitoring points, and then synthesizing each dimension to obtain the overall consistency index C between the ith monitoring point and the jth monitoring point ij and compared with the arbitration threshold Cth to determine whether the data of the ith monitoring point and the jth monitoring point are consistent. If not, it is determined as a single-point link failure, and the global alarm is not triggered. If consistent, it is determined as a global environmental anomaly, and the third strategy is given. Step five, through the whole process automatic generation of the link traceability log of the early warning instruction, the parameter collection, priority determination, window comparison and consistency arbitration steps are recorded completely, which is used for accident tracing and parameter optimization. 2.The distributed environmental parameter intelligent early warning method based on total control network streaming computation according to claim 1, wherein, The step one comprises: S11, divide the whole activity launch platform environment into several monitoring points, n monitoring points in the platform environment monitoring module according to the number Z1, Z2, Z3,..., Z n Sequential marking is carried out; the running process of the ith monitoring point is monitored in real time; the running data of the ith monitoring point is collected, including environmental temperature data, environmental humidity data, wind speed data and equipment running state data; S111, real-time monitoring of the umbilical tower base structure area and the hydraulic support area of the i th monitoring point of the active launch platform; installing high-temperature-resistant and vibration-resistant environmental sensors near the umbilical tower base, the hydraulic cylinder shell and the support arm; real-time collection and recording of environmental temperature parameters; S112, real-time monitoring of the electrical control cabin and the enclosed channel of the i th monitoring point of the active launch platform; arranging industrial humidity sensors in the air outlet of the electrical cabin, the cable trench and the sealed space; real-time collection and recording of environmental humidity parameters; S113, real-time monitoring of the gas guide and exhaust port, the platform contour and the air inlet of the walking system of the i th monitoring point of the active launch platform; installing three-dimensional ultrasonic wind speed and direction instruments in the guide and exhaust port, the platform contour air hole and the walking chassis air inlet; real-time collection of wind speed parameters; S114, real-time monitoring of the hydraulic system and power distribution area of the i th monitoring point of the active launch platform; installing current, voltage and vibration integrated sensors at the hydraulic pump station, the accumulator outlet, the switch cabinet and the key cable outlet; real-time collection of equipment operation state parameters, including current, voltage, hydraulic pressure and abnormal vibration signals. 3.The method of claim 2, wherein, The step one further comprises: S12, data preprocessing of the environmental temperature parameters, the environmental humidity parameters, the wind speed parameters and the electrical equipment operation state parameters; S121, using Kalman filtering and median filtering algorithm to suppress noise and remove abnormal points of temperature parameters and humidity parameters; removing high-frequency interference components of wind speed parameters through fast Fourier transform FFT analysis; using wavelet denoising method to eliminate pulse interference in current, voltage and hydraulic sampling of equipment operation state parameters; S122, the environmental temperature parameter, the environmental humidity parameter, the wind speed parameter and the equipment running state parameter are normalized by using a Z-score standardization method, and the dimensions are unified; a millisecond timestamp is stamped on each sampling data by using an NTP network time synchronization protocol, so that the time sequences of the multi-source data are consistent under a streaming calculation framework; S13, the processed environmental temperature parameter, the environmental humidity parameter, the wind speed parameter and the equipment running state parameter are transmitted to the platform general control network in an event-driven manner by establishing a parameter set of the ith monitoring point through an MQTT protocol and a Kafka streaming message queue.

4. The method according to claim 3, wherein, The step two includes: S21, the environmental temperature parameter normalized value Tg, the environmental humidity parameter normalized value Hd, the wind speed parameter normalized value Vs and the equipment running state parameter normalized value Ey in the parameter set of the ith monitoring point are extracted, and a priority coefficient Pyx is calculated and obtained.

5. The method according to claim 4, wherein, The step two further includes: S22, a first evaluation result is obtained by comparing and analyzing the priority coefficient Pyx and a preset priority threshold Pth through the priority threshold Pth. When the priority coefficient Pyx is greater than the priority threshold Pth, it indicates that the parameter stream of the current monitoring point meets the key safety parameter requirement, and automatically enters a high-priority channel, so that the current parameter stream is transmitted and processed within a second-level time delay. When the priority coefficient Pyx is less than or equal to the priority threshold Pth, it indicates that the parameter stream of the current monitoring point does not meet the key safety parameter requirement, a first warning instruction is triggered, and a first strategy is generated: the current parameter stream is placed in a common channel and transmitted in a batch manner or a long transmission cycle; a monitoring enhancement mechanism is started to increase the sampling frequency and data cache depth of the current monitoring point to prevent abnormal fluctuations from being ignored; if the parameter stream does not meet the key safety parameter requirement for three consecutive times, an artificial inspection prompt is triggered to prompt an operation and maintenance personnel to check the sensor state and data link.

6. The method according to claim 5, wherein, The step three includes: S31, a minute-level sliding window Wm and an hour-level sliding window Wh are respectively established under the streaming calculation framework of the ith monitoring point, segmented storage and real-time updating are performed on the environmental temperature, the environmental humidity, the wind speed and the equipment running state parameter, parameter sequence values XTm and XTh of the minute-level sliding window Wm and the hour-level sliding window Wh are captured at different time scales, and the parameter sequence values XTm and XTh of the minute-level sliding window Wm and the hour-level sliding window Wh are obtained.

7. The method according to claim 6, wherein, The step three further includes: S32, the parameter sequence values XTm and XTh in the minute-level sliding window Wm and the hour-level sliding window Wh are extracted, average values are respectively calculated, and a stability index SW is calculated after non-dimensional processing. S33, a second evaluation result is obtained by comparing and analyzing the stability index SW and a preset stability threshold Sth through the stability threshold Sth. When the stability index SW is less than or equal to the stability threshold Sth, it indicates that the matching degree of the current parameter sequence values XTm and XTh is qualified, and continuous monitoring is performed. When the stability index SW is greater than the stability threshold Sth, it indicates that the matching degree of the current parameter sequence value XTm and XTh is unqualified, there is a risk of short-term sudden change and trend deviation; the second early warning instruction is triggered, and the second strategy is generated: the current monitoring point is marked as a key data source, and the scheduling priority in the total control network is improved; at the same time, cross verification is performed on the short window and the long window data, and the accidental noise and single-point sensor distortion are eliminated; the sampling frequency and the data points are encrypted in the next two monitoring periods; if the stability index SW is greater than the stability threshold Sth for three consecutive monitoring periods, the consistency arbitration detection mechanism is started. 8.The method of claim 7, wherein, The step four comprises: S41, when the stability index SW meets the stability threshold Sth in three consecutive monitoring periods, start the consistency arbitration detection mechanism, extract the multi-dimensional environmental parameter data of the ith monitoring point and the adjacent jth monitoring point in the same time slice, adopt the vectorization representation method, construct into a multi-parameter vector form according to the parameter dimension, and get the node parameter vector set X i and X j . 9.The method of claim 8, wherein, The step four further comprises: S42, obtaining the consistency index C i and X j , using a normalization algorithm to standardize the parameter vectors dimension by dimension, eliminating the differences in the dimensions and orders of magnitude of different parameters, using a relative difference degree calculation method to calculate the consistency of the node parameters dimension by dimension, and obtaining the overall consistency index C ij between the i th monitoring point and the j th monitoring point by using a weighted average algorithm to synthesize the results of all dimensions. S43、through the preset arbitration threshold Cth, and the overall consistency index C ij comparing and analyzing with the arbitration threshold Cth, obtaining a third evaluation result includes: When the overall consistency index C ij When the arbitration threshold Cth, it is determined that the data of the ith monitoring point and the jth monitoring point are inconsistent, a single-point link failure is determined, a global alarm is not triggered, manual inspection or the use of a backup sensing node is notified. When the overall consistency index C ij When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C When the overall consistency index C 10. The method according to claim 9, wherein, The step five comprises: S51, through the whole process triggered by the first, second and third early warning instructions, a complete link tracing log is automatically generated, and each step from parameter collection, priority determination, window comparison and consistency arbitration is recorded in detail; the log is used as a basis for accident tracing and for parameter and model optimization correction.

Images