Mud pit detection device based on Internet of Things machine learning and control method thereof

Abstract

The invention relates to the technical field of mud pit detection, in particular to a mud pit detection device based on internet of things machine learning and a control method thereof.The control method comprises the steps that S1, the liquid level, the PH value, the temperature and the pressure parameters of a mud pit are obtained, and a weighted moving average filtering algorithm is used for preprocessing collected data; s2, uploading the processed data to a cloud platform by using a low-power-consumption NB-IoT module; s3, the cloud platform carries out storage, distributed processing and anomaly detection on the uploaded data, key parameters of the mud pit can be continuously monitored for 24 hours, the data accuracy is ensured, meanwhile, based on a machine learning model, the abnormal state is automatically recognized, early warning information is sent in advance, potential safety hazards are reduced, remote checking of the state of the mud pit is supported, and the system is convenient to use. The management process is simplified, the efficiency is improved, a mud pit state analysis report is provided, the mud treatment process is optimized, the maintenance cost is reduced, the mud pit state can be effectively monitored, leakage and overflow are prevented, and the construction area and the surrounding environment are protected.

Description

Technical Field

[0001] The present invention relates to the technical field of mud pool detection, and in particular to a mud pool detection device based on machine learning of the Internet of Things and a control method thereof. Background Art

[0002] In the field of civil engineering, especially in projects such as subway construction, tunneling and deep foundation pit excavation, the mud pool is a key component of the mud circulation system. The stability of its state is directly related to the construction progress, construction safety and protection of the surrounding environment.

[0003] However, traditional mud pool management mainly relies on manual inspection, which has problems such as long monitoring cycle, inaccurate data, and delayed response, and it is difficult to meet the needs of modern efficient and safe construction management. Therefore, a mud pool detection device and a control method based on Internet of Things machine learning are proposed to address the above problems. Summary of the invention

[0004] The purpose of the present invention is to provide a mud pool detection device and a control method based on machine learning of the Internet of Things, so as to solve the problem that traditional mud pool management mainly relies on manual inspections, has long monitoring cycles, inaccurate data, delayed response, etc., and is difficult to meet the needs of modern efficient and safe construction management.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A mud pool detection device and a control method thereof based on machine learning of the Internet of Things, comprising:

[0007] S1: Obtain the liquid level, pH value, temperature and pressure parameters of the mud pool, and use the weighted moving average filtering algorithm to pre-process the collected data;

[0008] S2: Use the low-power NB-IoT module to upload the processed data to the cloud platform;

[0009] S3: The cloud platform stores, distributes and detects anomalies in the uploaded data. The formula for determining whether the data is abnormal is:

[0010]

[0011] In the formula, x t is the current data, μ is the mean, σ is the standard deviation, and k is the threshold factor;

[0012] S4: Use the ridge regression model to predict the mud pool state and combine the key parameters collected in real time with historical data to perform state assessment;

[0013] S5: Based on the model prediction results and anomaly detection data, real-time warning information is sent to users when risks are detected;

[0014] S6: Store historical data to the cloud platform and generate trend analysis reports on the mud pool status to optimize the mud treatment process.

[0015] As a further optimization of the present invention, the process of the state evaluation includes the following steps:

[0016] S31: Input feature construction: The parameters such as the liquid level, temperature, pressure and pH value of the mud pool and its time series features are constructed as the input feature matrix X. The input matrix is ​​defined in the following form:

[0017]

[0018] In the formula, x ij represents the jth parameter at the i-th moment, n is the time step, and m is the number of features;

[0019] S32: Ridge regression prediction: Calculate the predicted value of the mud pool state based on the ridge regression model The prediction formula is:

[0020] in

[0021] Where W is the regression coefficient matrix, which describes the influence weight of each feature on the state, b is the bias term, and λ is the ridge regression regularization strength factor.

[0022] S33: Dynamic update and multi-objective prediction: Introduce sliding window technology to update the input matrix X and target variable matrix Y according to the new data collected in real time to achieve dynamic prediction;

[0023] The model supports multi-target prediction and is expanded into matrix form:

[0024]

[0025] In the formula, It is a multi-objective prediction matrix, which simultaneously outputs the liquid level change trend, overflow risk and pH abnormality possibility.

[0026] S34: According to the prediction results or Evaluate the changing trend of mud pool status, generate risk assessment reports, and issue early warning signals based on model prediction thresholds.

[0027] As a further optimized content of the present invention, the ridge regression regularization intensity factor satisfies the following adaptive calculation formula:

[0028]

[0029] Where η is a regularization adjustment parameter, which is used to adapt to different feature data distributions.

[0030] As a further optimization of the present invention, in the data collection and transmission step, the liquid level height is calculated by using a liquid level sensor through the following formula:

[0031]

[0032] Where H is the liquid level, P is the pressure, ρ is the mud density, and g is the gravitational acceleration.

[0033] As a further optimization of the present invention, in the data processing and analysis step, a sliding window algorithm is used to dynamically calculate the mean and standard deviation.

[0034] As a further optimization of the present invention, the machine learning model adopts a time series prediction model and introduces a time-related factor t into the prediction formula.

[0035] As a further optimization of the present invention, in the intelligent early warning step, the system sends multi-level alarm information to the user according to the risk level classification, including high risk, medium risk and low risk status.

[0036] As further optimized content of the present invention, it includes: a sensor array, a data acquisition and processing module, an Internet of Things module, a power module and a protective shell. The sensor array is used to collect key parameters such as mud pool level, pH value, temperature and pressure. The data acquisition and processing module is used to receive sensor signals and perform calibration and filtering. The Internet of Things module adopts a low-power NB-IoT communication protocol to achieve remote transmission of data. The power module provides stable power for the device, including batteries and solar panels. The protective shell adopts a waterproof, dustproof and explosion-proof design.

[0037] As further optimized content of the present invention, the liquid level sensor of the sensor array adopts a pressure measurement method and obtains the liquid level height through a liquid level calculation formula. The data acquisition and processing module adopts a weighted moving average filtering algorithm to smooth the collected signal and filter out random noise.

[0038] As further optimized content of the present invention, the Internet of Things module supports two-way communication function, the user can remotely send commands to adjust the sensor acquisition frequency, and the IP level of the protective shell reaches IP68.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. In the present invention, the key parameters of the mud pool can be monitored 24 hours a day to ensure data accuracy. At the same time, based on the machine learning model, abnormal conditions can be automatically identified, early warning information can be sent in advance, and potential safety hazards can be reduced. In addition, the mud pool status can be viewed remotely, the management process can be simplified, efficiency can be improved, and a mud pool status analysis report can be provided to optimize the mud treatment process and reduce maintenance costs. The mud pool status can be effectively monitored to prevent leakage and overflow, and to protect the construction area and surrounding environment.

[0041] 2. In the present invention, the key parameters such as liquid level, pH value, temperature, pressure, etc. are collected in real time through the sensor array, and the weighted moving average filtering algorithm and the sliding window anomaly detection model are combined to realize data smoothing and dynamic monitoring. At the same time, a multi-objective prediction model based on ridge regression is adopted, which combines real-time data with historical data to support multi-objective predictions such as liquid level trend, overflow risk, pH anomaly, etc. The model introduces adaptive regularization factors and time-related factors, optimizes feature weights, improves prediction accuracy and robustness, and the overall design realizes accurate evaluation of the mud pool state and reduces safety hazards;

[0042] 3. In the present invention, the device adopts a modular design, and the sensor array and the data acquisition module achieve efficient collaboration. The data is remotely transmitted to the cloud through a low-power NB-IoT communication module, supports two-way communication function, and can flexibly adjust the sensor acquisition frequency. The protective shell design reaches IP68 level, which can adapt to harsh environments such as high humidity, high dust and explosion hazards, and provide guarantee for the long-term reliable operation of the equipment. In addition, the system integrates a multi-level alarm mechanism to provide users with targeted alarm information according to the risk level, and combines the status analysis report generated by the cloud to optimize the mud treatment process and effectively reduce the operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a flow chart of a control method for a mud pool detection device based on machine learning of the Internet of Things in the present invention;

[0044] Figure 2 This is a system block diagram of a mud pool detection device based on Internet of Things machine learning in the present invention. DETAILED DESCRIPTION

[0045] See also Figure 1-2 , the present invention provides a technical solution:

[0046] A mud pool detection device and a control method thereof based on machine learning of the Internet of Things, comprising:

[0047] S1: Obtain the liquid level, pH value, temperature and pressure parameters of the mud pool, and use the weighted moving average filtering algorithm to pre-process the collected data;

[0048] S2: Use the low-power NB-IoT module to upload the processed data to the cloud platform;

[0049] S3: The cloud platform stores, distributes and detects anomalies in the uploaded data. The formula for determining whether the data is abnormal is:

[0050]

[0051] In the formula, x t is the current data, μ is the mean, σ is the standard deviation, and k is the threshold factor;

[0052] S4: Use the ridge regression model to predict the mud pool state and combine the key parameters collected in real time with historical data to perform state assessment;

[0053] S5: Based on the model prediction results and anomaly detection data, real-time warning information is sent to users when risks are detected;

[0054] S6: Store historical data to the cloud platform and generate trend analysis reports on the mud pool status to optimize the mud treatment process. Through real-time collection of key parameters, ridge regression model prediction, anomaly detection and intelligent early warning, it provides comprehensive mud pool status monitoring and risk assessment to achieve efficient and accurate safety management.

[0055] As a technical solution for further implementation of this plan, the process of status assessment includes the following steps:

[0056] S31: Input feature construction: The parameters such as the liquid level, temperature, pressure and pH value of the mud pool and its time series features are constructed as the input feature matrix X. The input matrix is ​​defined in the following form:

[0057]

[0058] In the formula, x ij represents the jth parameter at the i-th moment, n is the time step, and m is the number of features;

[0059] S32: Ridge regression prediction: Calculate the predicted value of the mud pool state based on the ridge regression model The prediction formula is:

[0060] in

[0061] Where W is the regression coefficient matrix, which describes the influence weight of each feature on the state, b is the bias term, and λ is the ridge regression regularization strength factor.

[0062] S33: Dynamic update and multi-objective prediction: Introduce sliding window technology to update the input matrix X and target variable matrix Y according to the new data collected in real time to achieve dynamic prediction;

[0063] The model supports multi-target prediction and is expanded into matrix form:

[0064]

[0065] In the formula, It is a multi-objective prediction matrix, which simultaneously outputs the liquid level change trend, overflow risk and pH abnormality possibility.

[0066] S34: According to the prediction results or Evaluate the changing trend of mud pool status, generate risk assessment reports, and issue warning signals based on the model prediction threshold. Through the dynamic prediction model of feature matrix construction and ridge regression, capture the time series characteristics of mud pool status, and achieve accurate prediction and trend analysis of multi-objective risks;

[0067] As a technical solution for further implementation of this solution, the ridge regression regularization intensity factor satisfies the following adaptive calculation formula:

[0068]

[0069] In the formula, η is a regularization adjustment parameter, which is used to adapt to different feature data distributions. An adaptive regularization strength factor is introduced to automatically adjust the generalization ability of the model according to the feature distribution, reduce overfitting, and improve prediction accuracy and model stability.

[0070] As a technical solution for further implementation of this scheme, in the data collection and transmission step, a liquid level sensor is used to calculate the liquid level height using the following formula:

[0071]

[0072] In the formula, H is the liquid level, P is the pressure, ρ is the mud density, and g is the gravity acceleration. The liquid level of the mud pool is accurately calculated by the liquid level sensor calculation formula and the pressure measurement method to ensure the reliability and real-time performance of liquid level monitoring.

[0073] As a technical solution for further implementation of this solution, in the data processing and analysis step, a sliding window algorithm is used to dynamically calculate the mean and standard deviation, and the sliding window algorithm dynamically updates the mean and standard deviation to improve the real-time performance and accuracy of anomaly detection and reduce the possibility of false positives and false negatives;

[0074] As a technical solution for further implementation of this plan, the machine learning model adopts a time series prediction model, introduces a time-related factor t into the prediction formula, and introduces a time-related factor into the machine learning prediction to improve the model's adaptability to complex dynamic environments and enhance the sensitivity of the prediction results to actual state changes;

[0075] As a technical solution for further implementation of this plan, in the intelligent early warning step, the system sends multi-level alarm information to users according to risk level classification, including high risk, medium risk and low risk status. Through the multi-level alarm mechanism, accurate early warning information is provided to users according to risk level, which optimizes emergency response and improves management efficiency;

[0076] As a technical solution for further implementation of this plan, it includes: sensor array, data acquisition and processing module, Internet of Things module, power module and protective shell. The sensor array is used to collect key parameters of mud pool level, pH value, temperature and pressure. The data acquisition and processing module is used to receive sensor signals and perform calibration and filtering. The Internet of Things module adopts low-power NB-IoT communication protocol to achieve remote transmission of data. The power module provides stable power for the device, including batteries and solar panels. The protective shell adopts waterproof, dustproof and explosion-proof design. The device structure design is complete. The combination of sensor array, efficient data processing module and durable protective shell provides reliable guarantee for long-term monitoring in complex environments.

[0077] As a technical solution for further implementation of this solution, the liquid level sensor of the sensor array adopts a pressure measurement method and obtains the liquid level height through a liquid level calculation formula. The data acquisition and processing module adopts a weighted moving average filtering algorithm to smooth the collected signal and filter out random noise. The pressure measurement method is combined with a weighted moving average filtering algorithm to ensure the smoothness and accuracy of the collected data and reduce the impact of environmental noise on the monitoring results.

[0078] As a technical solution for further implementation of this plan, the Internet of Things module supports two-way communication function. Users can remotely send commands to adjust the sensor acquisition frequency. The IP level of the protective casing reaches IP68, which is suitable for long-term use in high humidity, high dust and explosion-hazardous environments. The Internet of Things module supports two-way communication function, which improves the flexibility and remote operation capability of the device. At the same time, the high-level protective casing adapts to harsh environments and extends the service life of the equipment.

[0079] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. The above is only a preferred implementation of the present invention. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, ordinary technicians in this technical field can make several improvements, modifications or changes without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the inventive concept and technical solution to other occasions without improvement, should be regarded as the protection scope of the present invention.

Claims

1. A control method for a mud pool detection device based on machine learning of the Internet of Things, characterized in that: include: S1: Obtain the liquid level, pH value, temperature and pressure parameters of the mud pool, and use the weighted moving average filtering algorithm to pre-process the collected data; S2: Use the low-power NB-IoT module to upload the processed data to the cloud platform; S3: The cloud platform stores, distributes and detects anomalies of the uploaded data. The formula for determining whether the data is abnormal is: In the formula, x t is the current data, μ is the mean, σ is the standard deviation, and k is the threshold factor; S4: Use the ridge regression model to predict the mud pool state and combine the key parameters collected in real time with historical data to perform state assessment; S5: Based on the model prediction results and anomaly detection data, real-time warning information is sent to users when risks are detected; S6: Store historical data to the cloud platform and generate trend analysis reports on the mud pool status to optimize the mud treatment process.

2. The control method of a mud pool detection device based on machine learning of the Internet of Things according to claim 1 is characterized in that: The status assessment process includes the following steps: S31: Input feature construction: The parameters such as the liquid level, temperature, pressure and pH value of the mud pool and its time series features are constructed as the input feature matrix X. The input matrix is ​​defined in the following form: In the formula, x ij represents the jth parameter at the i-th moment, n is the time step, and m is the number of features; S32: Ridge regression prediction: Calculate the predicted value of the mud pool state based on the ridge regression model The prediction formula is: Where W is the regression coefficient matrix, which describes the influence weight of each feature on the state, b is the bias term, and λ is the ridge regression regularization strength factor. S33: Dynamic update and multi-objective prediction: Introduce sliding window technology to update the input matrix X and target variable matrix Y according to the new data collected in real time to achieve dynamic prediction; The model supports multi-target prediction and is expanded into matrix form: In the formula, It is a multi-objective prediction matrix, which simultaneously outputs the liquid level change trend, overflow risk and pH abnormality possibility. S34: According to the prediction results or Evaluate the changing trend of mud pool status, generate risk assessment reports, and issue early warning signals based on model prediction thresholds.

3. The control method of a mud pool detection device based on machine learning of the Internet of Things according to claim 2 is characterized in that: The ridge regression regularization strength factor satisfies the following adaptive calculation formula: Where η is a regularization adjustment parameter, which is used to adapt to different feature data distributions.

4. The control method of a mud pool detection device based on Internet of Things machine learning according to claim 1 is characterized in that: In the data collection and transmission step, the liquid level height is calculated using the following formula using a liquid level sensor: Where H is the liquid level, P is the pressure, ρ is the mud density, and g is the gravitational acceleration.

5. The control method of a mud pool detection device based on machine learning of the Internet of Things according to claim 1 is characterized in that: In the data processing and analysis steps, a sliding window algorithm is used to dynamically calculate the mean and standard deviation.

6. The control method of a mud pool detection device based on machine learning of the Internet of Things according to claim 1 is characterized in that: The machine learning model adopts a time series prediction model and introduces a time-related factor t into the prediction formula.

7. The control method of a mud pool detection device based on machine learning of the Internet of Things according to claim 1 is characterized in that: In the intelligent early warning step, the system sends multi-level alarm information to the user according to the risk level classification, including high risk, medium risk and low risk status.

8. A mud pool detection device based on Internet of Things machine learning according to any one of claims 1 to 7, characterized in that: include: A sensor array, a data acquisition and processing module, an Internet of Things module, a power module and a protective shell. The sensor array is used to collect key parameters such as mud pool level, pH value, temperature and pressure. The data acquisition and processing module is used to receive sensor signals and perform calibration and filtering. The Internet of Things module adopts the low-power NB-IoT communication protocol to achieve remote transmission of data. The power module provides stable power for the device, including batteries and solar panels. The protective shell adopts a waterproof, dustproof and explosion-proof design.

9. A mud pool detection device based on Internet of Things machine learning according to claim 8, characterized in that: The liquid level sensor of the sensor array adopts a pressure measurement method and obtains the liquid level height through a liquid level calculation formula. The data acquisition and processing module adopts a weighted moving average filtering algorithm to smooth the collected signal and filter out random noise.

10. The mud pool detection device based on Internet of Things machine learning according to claim 8, characterized in that: The IoT module supports two-way communication function, and the user can remotely send commands to adjust the sensor acquisition frequency. The IP level of the protective shell reaches IP68.

Images