Engineering truck multi-area environment control system based on AI

Abstract

The invention discloses an engineering vehicle multi-area environment control system based on AI, and particularly relates to the technical field of engineering vehicle environment control. Humidity, electric field, air pressure and dust multi-mode sensing arrays are arranged in multiple areas of an engineering vehicle, dust particle deposition distribution time sequence characteristics are extracted, and a deposition acceleration coefficient and a humidity response coefficient are calculated; calculating a dust deposition risk coefficient, comparing the dust deposition risk coefficient with a preset value, and marking a region exceeding a threshold value as a region in a dust deposition abnormal state; analyzing a deposition trend based on a time sequence prediction model, inputting environmental parameters, and predicting a dust deposition velocity curve of each area of the engineering vehicle in a future set time; according to the method, the static electricity, humidity, air pressure and dust coupling deposition risk is assessed in a prospective mode, a collaborative intervention strategy is triggered through a dynamic early warning threshold function, the effect of rapidly restraining the risk is achieved, and the problem that traditional single-parameter monitoring lags behind, and ventilation efficiency is suddenly reduced due to the fact that a coupling mechanism cannot be captured is effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering vehicle environmental control, and more specifically, to an AI-based multi-zone environmental control system for an engineering vehicle. Background Art

[0002] In dusty environments, such as open-pit mines and arid construction scenarios, construction vehicles continuously operate in environments with high dust concentration, low relative humidity and large fluctuations in electrostatic potential. Dust particles are easily deposited on the inner wall of the air duct, heat exchange fins and the surface of temperature and humidity / CO2 sensors under the combined action of electrostatic adsorption and humidity gradients. Conventional positive pressure air supply and single-stage filter elements can only intercept large-particle dust, and fine dust is repeatedly enriched with the circulating airflow, eventually leading to airflow blockage, reduced heat exchange efficiency and decreased cabin air quality.

[0003] The existing system lacks simultaneous monitoring and coupled analysis of humidity, potential and dust concentration, is unable to identify abnormal dust deposition triggered by the coupling effect, and has not established a targeted intervention mechanism. When deposition intensifies, it is difficult to restore ventilation efficiency in time by relying solely on passive cleaning or filter replacement. Although this can partially alleviate the problem, it cannot accurately match the deposition risk under complex working conditions, resulting in waste of resources and poor intervention effects. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an AI-based multi-zone environmental control system for engineering vehicles, which predicts deposition risks and performs active coordinated intervention of air pressure, humidity and static elimination to suppress deposition and sensor contamination caused by the superposition of static electricity, humidity and dust, so as to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions: an AI-based multi-zone environmental control system for engineering vehicles, comprising:

[0006] The environmental data acquisition module collects environmental parameters such as humidity, electric field strength, air pressure, and dust concentration data in each area of ​​the engineering vehicle by installing humidity sensors, electrostatic field strength meters, air pressure sensors, and dust sensors in each area of ​​the engineering vehicle, and records the distribution map of dust particle deposition in each area of ​​the engineering vehicle;

[0007] The dust deposition anomaly identification module extracts the time series characteristics of the dust particle deposition distribution map, calculates the deposition acceleration coefficient and humidity response coefficient, and calculates the dust deposition risk factor based on the deposition acceleration coefficient and humidity response coefficient. It compares the dust particle deposition distribution time series characteristics with the preset value and marks the area exceeding the threshold as an area of ​​dust deposition anomaly.

[0008] The dust deposition risk prediction module analyzes deposition trends in areas experiencing abnormal dust deposition based on a time series prediction model. By inputting environmental parameters, it predicts the dust deposition velocity curve for each area of ​​the engineering vehicle within a set timeframe. Input features of the time series prediction model include humidity change rate, potential fluctuation trend, deposition velocity change rate, deposition amount trend, cross-regional airflow characteristics, and time lag characteristics.

[0009] The dust deposition intervention module, based on the predicted dust deposition rate curve, uses a multi-objective optimization algorithm to dynamically calculate the coordinated intervention strategy of air pressure regulation, humidity regulation and electrostatic elimination to effectively reduce the rate of dust deposition. Among them, air pressure regulation prioritizes reducing the pressure difference in the cross-regional airflow characteristics to suppress dust migration, humidity regulation adaptively controls the dew point difference based on the condensation risk probability to reduce sensor drift, and electrostatic elimination adjusts the discharge frequency according to the potential fluctuation trend to weaken particle adsorption. At the same time, the monitoring and adjustment module monitors the condensation status of the humidity sensor surface in real time, compensates and corrects the humidity data affected by condensation, and dynamically adjusts the intervention strategy.

[0010] Preferably, the system includes a region division module, which divides the engineering vehicle into regions based on the airflow path, equipment layout and dust deposition characteristics in the engineering vehicle; the dust deposition characteristics are obtained by analyzing the air duct path, airflow direction and flow distribution of the airflow path, and K-means clustering is used to partition the engineering vehicle cabin.

[0011] Preferably, the deposition acceleration coefficient is obtained by calculating the rate of change of dust concentration and air pressure; collecting dust concentration data and air pressure data in each area, calculating the rate of change of dust concentration, that is, the amount of change in dust concentration per unit time; calculating the rate of change of air pressure, that is, the amount of change in air pressure per unit time; multiplying the rate of change of air pressure by the rate of change of dust concentration to obtain the deposition acceleration coefficient, thereby indicating the acceleration trend of dust deposition; the deposition acceleration coefficient reflects the impact of environmental changes on dust deposition, in particular, the promoting or slowing effect of air pressure changes on the dust deposition rate;

[0012] The humidity response coefficient is obtained by calculating the humidity change rate and the average humidity level of the airflow; the humidity change in the area is monitored in real time, the rate of humidity change is calculated, and the average humidity level of the area is calculated; after dimensionless processing, the humidity change rate and the average humidity level of the area are weighted to obtain the humidity response coefficient. The humidity response coefficient reflects the impact of humidity changes on dust deposition. The greater the humidity change, the more obvious the impact on the deposition process. Therefore, this coefficient can quantify the impact of humidity on the deposition rate.

[0013] The dust deposition risk factor is obtained by multiplying the deposition acceleration factor and the humidity response factor. The deposition acceleration factor reflects the acceleration of dust deposition, while the humidity response factor reflects the impact of humidity changes on deposition. The dust deposition risk factor comprehensively considers the acceleration trend of dust deposition and the impact of humidity on the deposition process to provide a comprehensive risk assessment result. When the dust deposition risk factor exceeds the set threshold, the system will trigger appropriate intervention measures to reduce the risk of dust deposition and perform environmental adjustments.

[0014] Preferably, the system includes an early warning module connected to the dust deposition risk prediction module, which compares the predicted dust deposition rate curve with the early warning threshold. When the early warning threshold is exceeded, an early warning is issued and the dust deposition intervention module is executed; a response window is defined based on the response time of the intervention strategy and the confidence interval of the historical deposition rate data, and the dust deposition amount within the response window is calculated by a time series prediction model; the static basic early warning threshold is associated with the rate of change of the predicted deposition rate to generate a dynamic early warning threshold function; the early warning threshold is tightened within the response window as the dust deposition rate increases; the early warning threshold is scaled and corrected according to the real-time humidity difference and the electric field strength deviation to suppress the false high warning threshold in high humidity or strong electric field environments; when the cumulative deposition amount reaches a preset proportion or the operating cycle exceeds the limit, the dynamic early warning threshold function is recalibrated to eliminate the influence of sensor drift and environmental aging.

[0015] Preferably, the method for acquiring the time series prediction model includes the following steps:

[0016] Step S1: Data acquisition and preprocessing: Acquire real-time monitoring time series data of humidity values, electric field strength, dust concentration, current dust deposition rate, and dust deposition amount in each area of ​​the engineering vehicle; perform synchronous preprocessing through Z-score normalization and sliding window denoising to form a standardized data sequence to be analyzed;

[0017] Step S2: Feature extraction: extract the features of the data sequence to be analyzed and output a feature vector, including: humidity change rate; potential fluctuation trend; deposition rate change rate, deposition amount trend, cross-regional airflow characteristics, and time lag characteristics;

[0018] Step S3: Model construction: Taking the feature vector as input, a graph convolution-bidirectional LSTM hybrid network is adopted, in which the graph convolution layer extracts spatial associations based on the pre-calculated airflow adjacency matrix. The pre-calculated airflow adjacency matrix is ​​generated through offline CFD simulation and covers the airflow distribution under typical working conditions. In real-time operation, the adjacency matrix weights are dynamically adjusted to reflect airflow changes through online interpolation methods (such as weighted interpolation based on pressure sensor and wind speed sensor data). The bidirectional LSTM captures time series dependencies and adaptively adjusts the adjacency matrix weights in each cycle to reflect turbulence evolution. The dust deposition velocity curve of each area of ​​the engineering vehicle within a set time in the future is predicted. The prediction error of the dust deposition velocity curve is used as the loss function, and the model parameters are optimized through cross-validation. The output is the dust deposition velocity curve of each area, which represents the trend of the deposition velocity per unit time over time.

[0019] Step S4: Model verification and calibration: Based on historical operating data and real-time measurement data, the accuracy of the predicted dust deposition velocity curve is verified; when the deviation between the predicted value and the actual measurement value exceeds a preset value, the adaptive calibration procedure is triggered.

[0020] Preferably, the method for obtaining the time-lag feature comprises the following steps:

[0021] The nonlinear correlations between humidity, potential, deposition velocity, and dust deposition were calculated. The time-lag relationships between each parameter and dust deposition were modeled using mutual information metrics and nonlinear cross-correlation analysis, generating eigenvectors describing the nonlinear time lag between each parameter.

[0022] A neural network regression model is used, which includes an input layer, one or more hidden layers, and an output layer. The input layer receives humidity, potential, air pressure difference, and deposition rate parameters, the hidden layer uses a fully connected layer, and the output layer predicts the time-lag relationship between humidity, potential, and air pressure difference parameters and dust deposition. Training is performed using a backpropagation algorithm, with the loss function being the mean square error, and optimization is performed using a gradient descent method.

[0023] Based on real-time monitoring data, the model parameters of the time-lag characteristics are adjusted through an adaptive algorithm. The adaptive algorithm compares the actual monitoring results with the model prediction results at each feedback and adjusts the network weights to ensure the accuracy of the time-lag characteristics.

[0024] Based on the airflow adjacency matrix generated by CFD simulation, the spatial topological relationship of the airflow in the adjacency matrix is ​​used as input and combined with the fluid dynamics model to optimize the calculation of time-delay characteristics. The accuracy of nonlinear time-delay modeling is improved by optimizing the input characteristics.

[0025] The calculated time-lag characteristics are fed back to the time series prediction model. When the predicted deposition rate and the actual measurement data exceed the preset error, the adaptive calibration program is triggered to adjust the neural network regression model through a retraining process.

[0026] Preferably, by deploying miniature airflow velocity sensors at the corners of air ducts, narrow spaces or airflow intersections, the velocity gradient and turbulence characteristics of the local micro-airflow are monitored in real time; after the micro-airflow disturbance data are fused, the local micro-airflow disturbance intensity is generated, including the velocity gradient and turbulence characteristics. When the monitoring value exceeds the threshold, the adjustable air valve at the corresponding position or the auxiliary small fan is controlled to adjust the airflow direction or flow rate to weaken the turbulence or enhance the scouring effect and reduce dust deposition.

[0027] Preferably, in order to solve the problem that the humidity sensor causes data drift due to local condensation and thus misjudges the abnormal state of dust deposition, the dew point temperature is calculated in real time. When the difference between the ambient temperature and the dew point exceeds the preset value, the dew point compensation correction program is triggered to start, and the humidity measurement value affected by condensation is compensated and corrected in real time, thereby eliminating the interference of the humidity sensor drift on the prediction result and improving the accuracy of the judgment of the abnormal state of dust deposition.

[0028] The explanation states that the system collects environmental parameters such as humidity, electric field strength, air pressure and dust concentration in real time through a multimodal sensor array, and combines it with a graph convolution-bidirectional LSTM hybrid network to accurately predict the dust deposition rate curve, capture the complex coupling effects of static electricity, humidity, air pressure and dust, and avoid misjudgment of deposition risks caused by lag in traditional single-parameter monitoring. The graph convolution layer extracts spatial correlations between regions based on the airflow adjacency matrix generated by offline CFD simulation, and dynamically adjusts weights online through interpolation of air pressure and wind speed sensor data to reflect real-time turbulence changes. The bidirectional LSTM mines long-term dependencies in time series to ensure the adaptability of the prediction model to dynamic working conditions. The dynamic intervention strategy uses a multi-objective optimization algorithm to comprehensively consider the interactive constraints of air pressure regulation, humidity control and static electricity elimination. , by prioritizing the reduction of cross-regional air pressure differences to suppress dust migration, adaptively adjusting the dew point difference to reduce the drift of the humidity sensor caused by condensation, and optimizing the static elimination frequency according to the potential fluctuation trend, weakening the particle adsorption effect, thereby effectively reducing the amount of dust deposition on the inner wall of the air duct and the heat exchange fins; the nonlinear coupling calibration mechanism further integrates the characteristics of charge heterogeneity, local micro-airflow disturbance intensity, condensation risk and oil mist composite ratio, and dynamically corrects the deposition velocity curve in the form of exponential decay integral to match the inertia changes of the operating conditions of engineering vehicles, ensuring the prediction accuracy and robustness of the intervention effect; logically forming a complete chain from multi-source perception to accurate prediction to collaborative intervention, significantly improving ventilation efficiency, sensor reliability and system maintenance economy, and reasonably solving the deposition problem in complex dust environments.

[0029] Preferably, the multi-objective optimization algorithm includes:

[0030] Air pressure regulation targets, which adjusts air pressure in real time to optimize airflow and prevent accelerated dust deposition by minimizing the deposition acceleration factor;

[0031] Humidity adjustment target, by calculating the humidity response coefficient, adjusts the humidity range in real time to prevent excessive humidity fluctuations from causing abnormal dust deposition;

[0032] Static elimination goal: Based on a comprehensive model of electric field strength and humidity changes, optimize the static elimination strategy to reduce the impact of static electricity accumulation on dust deposition;

[0033] The air pressure regulation target, humidity regulation target and static elimination target are integrated through a weighted objective function, wherein the weights of the targets are dynamically adjusted according to real-time environmental data and prediction results, thereby achieving coordinated optimization of air pressure, humidity and static elimination.

[0034] Technical effects and advantages of the present invention:

[0035] (1) The present invention provides a solution that deploys humidity, electric field, air pressure and dust multimodal sensor arrays in multiple areas of the engineering vehicle, and combines them with a time series prediction model to generate a dust deposition rate curve in real time, thereby achieving a forward-looking assessment of the risk of electrostatic-humidity-dust coupled deposition, and triggering a collaborative intervention strategy through a dynamic early warning threshold function to achieve the effect of rapid risk containment, effectively solving the problem of traditional single-parameter monitoring lag and inability to capture the coupling mechanism, resulting in a sudden drop in ventilation efficiency.

[0036] (2) The present invention provides a solution that calculates the coordinated intervention curve of air pressure, humidity and static elimination through multi-objective optimization, thereby achieving adaptive reduction of pressure differential scour and condensation deviation without affecting the cabin comfort, and continuously updates the time series prediction model with the help of drift detection and nonlinear coupling calibration mechanism, effectively solving the problems of short filter life, large sensor drift and high maintenance cost under high dust conditions, while significantly reducing energy consumption and improving system reliability and maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a simplified structural diagram of the AI-based multi-zone environmental control system for engineering vehicles of the present invention. DETAILED DESCRIPTION

[0038] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0039] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0040] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0041] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0042] Example 1, see Figure 1 The present invention provides a simplified structure diagram of the multi-zone environmental control system for engineering vehicles based on AI. Figure 1 The AI-based multi-zone environmental control system for engineering vehicles shown includes:

[0043] The environmental data acquisition module collects environmental parameters in each area of ​​the engineering vehicle, including humidity, electric field strength, air pressure, and dust concentration data, by installing humidity sensors, electrostatic field strength meters, air pressure sensors, and dust concentration sensors in each area. It also uses high-speed photography or electrostatic imaging to record the distribution of dust particle deposition in each area of ​​the engineering vehicle.

[0044] The dust deposition anomaly identification module extracts the time series characteristics of the dust particle deposition distribution map, calculates the deposition acceleration coefficient and humidity response coefficient, and calculates the dust deposition risk factor based on the deposition acceleration coefficient and humidity response coefficient. It compares the coefficient with the preset value and marks the area exceeding the threshold as an area with abnormal dust deposition status.

[0045] The dust deposition risk prediction module analyzes deposition trends in areas experiencing abnormal dust deposition based on a time series prediction model. By inputting environmental parameters, it predicts the dust deposition velocity curve for each area of ​​the engineering vehicle within a set timeframe. Input features of the time series prediction model include humidity change rate, potential fluctuation trend, deposition velocity change rate, deposition amount trend, cross-regional airflow characteristics, and time lag characteristics.

[0046] The dust deposition intervention module, based on the predicted dust deposition rate curve, uses a multi-objective optimization algorithm to dynamically calculate the coordinated intervention strategy of air pressure regulation, humidity regulation and electrostatic elimination to effectively reduce the rate of dust deposition. Among them, air pressure regulation prioritizes reducing the pressure difference in the cross-regional airflow characteristics to suppress dust migration, humidity regulation adaptively controls the dew point difference based on the condensation risk probability to reduce sensor drift, and electrostatic elimination adjusts the discharge frequency according to the potential fluctuation trend to weaken particle adsorption. At the same time, the monitoring and adjustment module monitors the condensation status of the humidity sensor surface in real time, compensates and corrects the humidity data affected by condensation, and dynamically adjusts the intervention strategy.

[0047] In a possible embodiment, the multi-objective optimization algorithm includes:

[0048] Air pressure regulation targets, which adjusts air pressure in real time to optimize airflow and prevent accelerated dust deposition by minimizing the deposition acceleration factor;

[0049] Humidity adjustment target, by calculating the humidity response coefficient, adjusts the humidity range in real time to prevent excessive humidity fluctuations from causing abnormal dust deposition;

[0050] Static elimination goal: Based on a comprehensive model of electric field strength and humidity changes, optimize the static elimination strategy to reduce the impact of static electricity accumulation on dust deposition;

[0051] The air pressure regulation target, humidity regulation target and static elimination target are integrated through a weighted objective function, wherein the weights of the targets are dynamically adjusted according to real-time environmental data and prediction results, thereby achieving coordinated optimization of air pressure, humidity and static elimination.

[0052] In this embodiment, the multi-objective optimization algorithm simultaneously considers the interaction between the three when adjusting air pressure, humidity and static elimination; changes in air pressure affect airflow velocity and dust distribution, changes in humidity affect static elimination efficiency, and static adjustment will also change the ambient humidity; the multi-objective optimization algorithm uses humidity, electric field and air pressure sensor data as input, and calculates the optimal value of each parameter at the same time; when the humidity is too high, the humidity is reduced and static elimination is enhanced; when the electrostatic field is too high, static elimination is enhanced and the air pressure is adjusted to promote airflow. Through the above-mentioned parameter association settings, this embodiment realizes the coordinated control of the three intervention measures.

[0053] Explanation: With the output intervention strategy as the control input, the humidity regulator (such as a vehicle-mounted dehumidifier or ultrasonic humidifier) ​​is dynamically adjusted to precisely control the regional humidity, and the ion wind generator or ultrasonic atomizing nozzle installed in the air duct and key positions is activated to actively eliminate static charges in the air and on surfaces; at the same time, the static neutralization rod is started to eliminate the surface charge of the bulkhead and sensor to avoid the electrostatic adsorption effect of dust; during the execution process, the humidity, potential, and dust sensor feedback data are continuously monitored and compared with the predicted target value in real time to form a complete closed-loop control.

[0054] To achieve the linkage between the prediction model and feedback control, this embodiment continuously compares the measured deposition rate with the predicted target in real time, and the difference between the two is used as a feedback signal to drive the dynamic adjustment of the execution strategy; when the measured deposition rate is higher than the predicted value, the ion wind intensity can be increased or the static elimination time can be extended; if it is lower than the prediction, the control intensity is reduced accordingly; the fine-tuning process can use proportional adjustment or incremental correction algorithm to dynamically update parameters according to the size of the deviation. Through this closed-loop feedback, this embodiment can adaptively bring the actual state closer to the predicted target.

[0055] In one possible embodiment, the system includes a region division module that divides the engineering vehicle into regions based on the airflow path, equipment layout, and dust deposition characteristics within the engineering vehicle; dust deposition characteristics are obtained by analyzing the air duct path, airflow direction, and flow distribution of the airflow path, and K-means clustering is used to partition the engineering vehicle cabin;

[0056] In one possible embodiment, the regions are divided based on the physical structure (such as equipment layout and ventilation system partitioning), and the boundaries are optimized using a clustering algorithm to ensure that the division results are both consistent with the actual structure and have data-driven characteristics; the initial regions are defined by compartments and ventilation duct partitions, and the fuzzy boundaries (such as equipment gaps) are refined using a clustering algorithm.

[0057] In one possible embodiment, the deposition acceleration coefficient is obtained by calculating the rate of change of dust concentration and air pressure. Dust concentration data and air pressure data in each area are collected to calculate the rate of change of dust concentration, that is, the amount of change in dust concentration per unit time; the rate of change of air pressure, that is, the amount of change in air pressure per unit time, is calculated; the rate of change of air pressure is multiplied by the rate of change of dust concentration to obtain the deposition acceleration coefficient, thereby indicating the acceleration trend of dust deposition. The deposition acceleration coefficient reflects the impact of environmental changes on dust deposition, particularly the effect of air pressure changes on the acceleration or reduction of dust deposition rate.

[0058] The humidity response coefficient is obtained by calculating the humidity change rate and the average humidity level of the airflow; the humidity change in the area is monitored in real time, the rate of humidity change is calculated, and the average humidity level of the area is calculated; after dimensionless processing, the humidity change rate and the average humidity level of the area are weighted to obtain the humidity response coefficient. The humidity response coefficient reflects the impact of humidity changes on dust deposition. The greater the humidity change, the more obvious the impact on the deposition process. Therefore, this coefficient can quantify the impact of humidity on the deposition rate.

[0059] The dust deposition risk factor is obtained by multiplying the deposition acceleration factor and the humidity response factor. The deposition acceleration factor reflects the acceleration of dust deposition, while the humidity response factor reflects the impact of humidity changes on deposition. The dust deposition risk factor comprehensively considers the acceleration trend of dust deposition and the impact of humidity on the deposition process to provide a comprehensive risk assessment result. When the dust deposition risk factor exceeds the set threshold, the system will trigger appropriate intervention measures to reduce the risk of dust deposition and perform environmental adjustments.

[0060] In one possible embodiment, the system includes an early warning module connected to a dust deposition risk prediction module, which compares the predicted dust deposition rate curve with the early warning threshold. When the early warning threshold is exceeded, an early warning is issued and a dust deposition intervention module is executed; a response window is defined based on the response time of the intervention strategy and the confidence interval of the historical deposition rate data, and the dust deposition amount within the response window is calculated through a time series prediction model; a static basic early warning threshold is associated with the rate of change of the predicted deposition rate to generate a dynamic early warning threshold function; the early warning threshold is tightened within the response window as the dust deposition rate increases; the early warning threshold is scaled and corrected according to the real-time humidity difference and the electric field strength deviation to suppress the falsely high early warning threshold in high humidity or strong electric field environments; when the cumulative deposition amount reaches a preset proportion or the operating cycle exceeds the limit, the dynamic early warning threshold function is recalibrated to eliminate the effects of sensor drift and environmental aging.

[0061] In a possible embodiment, the cumulative deposition amount is calculated by integrating the deposition rate over time; and the abnormal state is determined based on the cumulative deposition amount and a corresponding threshold.

[0062] In a possible embodiment, the dynamic warning threshold function represents the dust deposition rate warning threshold that changes with time. The dynamic warning threshold function T dynamic The formula for (t) is:

[0063]

[0064] Among them, T base It is a static basic warning threshold, determined based on the upper limit of the 95% confidence interval of historical sedimentation rate data; To predict the deposition rate change rate, a time series prediction model is used for calculation; ΔH(t) is the real-time humidity difference; △E(t) is the electric field intensity deviation; k1 is the change rate sensitivity coefficient, which reflects the amplification effect of the deposition rate change on the threshold; k2 and k3 are the linear scaling coefficients of the humidity difference and the electric field intensity deviation, respectively, which adjust the inhibitory effect of environmental conditions on the value.

[0065] It should be further explained in the embodiment of the present invention that the method for obtaining the time series prediction model includes the following steps:

[0066] Step S1: Data acquisition and preprocessing: Acquire real-time monitoring time series data of humidity values, electric field strength, dust concentration, current dust deposition rate, and dust deposition amount in each area of ​​the engineering vehicle, and output standardized data to be analyzed after preprocessing;

[0067] The specific implementation method is as follows: heterogeneous data are collected on the experimental platform and the engineering vehicle site, including humidity sensor data, surface electrostatic field strength meter data, air pressure sensor data, and dust concentration sensor data at different locations, and the distribution of local particle deposition locations is recorded; the heterogeneous data are denoised and interpolated using the sliding window method (synchronous preprocessing is performed through Z-score normalization and sliding window denoising);

[0068] Step S2: Feature extraction: Extract the features of the data sequence to be analyzed and output a feature vector, including: humidity change rate (difference in relative humidity per unit time); potential fluctuation trend (based on the time series variance of the electrostatic potential); deposition rate change rate, deposition amount trend, cross-regional airflow characteristics (airflow exchange rate calculated by differential pressure sensor and wind speed sensor), and time lag characteristics (cross-correlation analysis is used to calculate the correlation coefficients of humidity, potential, and deposition rate with dust deposition).

[0069] Step S3: Model construction: Taking the feature vector as input, a graph convolution-bidirectional LSTM hybrid network is adopted, in which the graph convolution layer extracts spatial associations based on the pre-calculated airflow adjacency matrix. The pre-calculated airflow adjacency matrix is ​​generated through offline CFD simulation and covers the airflow distribution under typical working conditions. In real-time operation, the adjacency matrix weights are dynamically adjusted to reflect airflow changes through online interpolation methods (such as weighted interpolation based on pressure sensor and wind speed sensor data). The bidirectional LSTM captures time series dependencies and adaptively adjusts the adjacency matrix weights in each cycle to reflect turbulence evolution. The dust deposition velocity curve of each area of ​​the engineering vehicle within a set time in the future (for example, 1 hour) is predicted. The dust deposition velocity curve prediction error is used as the loss function, and the model parameters are optimized through cross-validation (the number of hidden layer units is 64, the learning rate is 0.001, and the number of training rounds is 100). The output is the dust deposition velocity curve of each area, which represents the trend of the deposition velocity per unit time over time.

[0070] Step S4: Model verification and calibration: Based on historical operating data and real-time measurement data (calibrated by sediment sensors or optical dust sensors), the accuracy of the predicted dust deposition rate curve is verified; when the deviation between the predicted value and the actual measured value exceeds a preset value, such as 5%, the adaptive calibration procedure is triggered.

[0071] In this embodiment, spatial correlation features are constructed through the airflow adjacency matrix (using CFD simulation or sensor data) and extracted by the graph convolution layer; the time series correlation features are obtained through cross-correlation analysis to obtain the influence weights of humidity, potential and deposition rate and constitute the input vector; the spatial features output by the graph convolution are input into the LSTM together with the time series to capture the time dependency; the graph convolution layer and the LSTM layer interact through feature transfer: the spatial features guide the time series prediction, and the time series output can be used to adjust the graph convolution weights, thereby completing the joint modeling of spatiotemporal features.

[0072] 10. It should be further explained in the embodiment of the present invention that the method for obtaining the time lag feature includes the following steps:

[0073] The nonlinear correlations between humidity, potential, deposition velocity, and dust deposition were calculated. The time-lag relationships between each parameter and dust deposition were modeled using mutual information metrics and nonlinear cross-correlation analysis, generating eigenvectors describing the nonlinear time lag between each parameter.

[0074] It is explained that during the implementation process, choosing a suitable mutual information measurement method, such as the mutual information measurement based on Shannon entropy, can effectively evaluate the nonlinear relationship between parameters such as humidity, potential, deposition rate and dust deposition; using a nonlinear regression model (such as support vector regression) to model the time-lag relationship between each parameter and dust deposition, using adaptive window length and step size to capture the nonlinear time-lag effect at different time scales, and analyzing the sensitivity of different time windows and parameters in the experiment, adjusting the parameters in the correlation calculation, and optimizing the match with the actual dust deposition scene, so as to more accurately describe the nonlinear coupling characteristics.

[0075] A neural network regression model is used, which includes an input layer, one or more hidden layers, and an output layer. The input layer receives humidity, potential, air pressure difference, and deposition rate parameters, the hidden layer uses a fully connected layer, and the output layer predicts the time-lag relationship between humidity, potential, and air pressure difference parameters and dust deposition. Training is performed using a backpropagation algorithm, with the loss function being the mean square error, and optimization is performed using a gradient descent method.

[0076] Explanation: When designing the neural network architecture, each input feature is standardized according to its dimension and unit; 2 to 3 hidden layers are selected, and the number of neurons in each layer is adjusted according to the actual data set size, usually 128 to 512 neurons. The depth of the network and the number of neurons in each layer are determined through cross-validation to prevent overfitting; during the training process, the network weights are optimized through back propagation and mean square error loss function; the training data set and validation set are split through time series to ensure the representativeness and diversity of the data and avoid the model's dependence on a specific environmental condition.

[0077] Based on real-time monitoring data, the model parameters of the time-lag characteristics are adjusted through an adaptive algorithm. The adaptive algorithm compares the actual monitoring results with the model prediction results at each feedback and adjusts the network weights to ensure the accuracy of the time-lag characteristics.

[0078] The adaptive algorithm uses an incremental learning strategy based on real-time feedback. It dynamically adjusts network weights by calculating the error between model predictions and actual monitoring results. This involves using the prediction error as input during each model update, adjusting the learning rate, and flexibly adjusting model parameters based on the magnitude of the error. This feedback mechanism ensures the model can adapt to changing environments, updating time-lag characteristics in real time and improving prediction accuracy.

[0079] Based on the airflow adjacency matrix generated by CFD simulation, the spatial topological relationship of the airflow in the adjacency matrix is ​​used as input and combined with the fluid dynamics model to optimize the calculation of time-delay characteristics. The accuracy of nonlinear time-delay modeling is improved by optimizing the input characteristics.

[0080] Explanation: Detailed airflow characteristics (including velocity distribution, vortex, etc.) are provided by CFD simulation results and integrated with time-lag characteristics; CFD simulation results are combined with real-time sensor data (such as airflow velocity, pressure, etc.) through a weighted fusion algorithm to optimize the calculation of time-lag characteristics; Kalman filtering and other methods are used to reduce the impact of simulation errors, thereby ensuring that the airflow model matches the actual situation of dust deposition and improving prediction accuracy.

[0081] The calculated time-lag characteristics are fed back to the time series prediction model. When the predicted deposition rate and the actual measured data exceed the preset error, the adaptive calibration program is triggered to adjust the neural network regression model through a retraining process.

[0082] Explanation: The calculated time-lag features are fused with the historical data and prediction results in the time series prediction model (for example, through weighted fusion methods and fusion with other input features in the time series prediction model) as one of the input features; the feedback mechanism quantifies the contribution of the time-lag features to the sedimentation velocity prediction by adjusting the time step, and dynamically updates the model input by combining historical data and real-time sensor data; through recurrent neural networks (RNN) or long short-term memory networks (LSTM), the prediction weights of the model are adjusted according to the time-lag features in each prediction cycle to ensure that their impact on the sedimentation velocity prediction is accurately captured.

[0083] In a possible embodiment, cross-scale time-lag features are extracted by multi-scale wavelet coherence-improved grey mutual information joint analysis, and main time-lag feature components sensitive to sedimentation rate are automatically screened by Shapley value sorting.

[0084] In one possible embodiment, a physical information regularization network is used to learn the coupling coefficient online, and the coupling weight is iteratively updated every time a 1-hour sampling window is completed to avoid inaccuracy of the fixed empirical formula.

[0085] Explanation: The following is a detailed explanation of local features and how to obtain them:

[0086] Particle charge distribution reflects the differences in charge distribution of dust particles in space due to electrostatic effects, which directly affects deposition behavior. This is obtained by collecting multi-point potential data using a surface electrostatic field meter (range ±5kV) and calculating the standard deviation and spatial distribution of the charge distribution.

[0087] The local micro-airflow disturbance intensity is used to describe the velocity changes and turbulence characteristics of the local airflow, which affects the suspension and deposition paths of dust. This is obtained by collecting velocity data using micro-airflow velocity sensors (deployed at key locations such as duct corners) and calculating the velocity gradient (rate of velocity change per unit time) and turbulence intensity (based on spatial spectrum analysis).

[0088] Condensation risk indicates the tendency for condensation to occur due to abnormal humidity fluctuations, which affects sensor data and deposition rate. This risk is acquired by analyzing the time series of humidity sensor data, detecting abnormal peaks or fluctuation frequencies in the humidity curve, and calculating the condensation probability based on temperature data.

[0089] The oil mist composite ratio reflects the deposition characteristics and adhesion ability of composite particles formed after the dust and oil mist are mixed. It is obtained by calculating the deposition slope (the rate of change of concentration per unit time) and adhesion strength (based on the ratio of the deposition amount to the residual amount after airflow flushing) through the dust concentration sensor and humidity data.

[0090] In one possible embodiment, in order to solve the problem of abnormal dust deposition caused by local micro-airflow disturbances in the cabin of an engineering vehicle, micro-airflow velocity sensors are deployed at the corners of air ducts, narrow spaces, or airflow intersections to monitor the velocity gradient and turbulence characteristics of the local micro-airflow in real time. After the micro-airflow disturbance data are fused, the local micro-airflow disturbance intensity is generated, including the velocity gradient (reflecting the spatial rate of change of the airflow velocity) and the turbulence characteristics (the turbulence intensity is quantified through spatial spectrum analysis). When the monitored value exceeds the threshold, the adjustable air valve at the corresponding position or the auxiliary small fan is controlled to adjust the airflow direction or flow rate to weaken the turbulence or enhance the scouring effect, thereby reducing dust deposition.

[0091] Example 2: The difference between this embodiment of the present invention and Example 1 is as follows:

[0092] In one possible embodiment, to address the issue of humidity sensor data drift due to local condensation, which can lead to misjudgment of abnormal dust deposition conditions, the dew point temperature is calculated in real time. When the difference between the ambient temperature and the dew point exceeds a preset value (e.g., a dew point difference of ≤ 2°C), a dew point compensation program is triggered to perform real-time compensation correction on humidity measurements affected by condensation. This eliminates the interference of humidity sensor drift on prediction results and improves the accuracy of abnormal dust deposition condition determination.

[0093] The specific implementation is as follows: the steps of compensating and correcting humidity data affected by condensation include the following steps:

[0094] Step S21: arranging a condensation monitoring electrode on the surface of the humidity sensor and monitoring the change of resistance or capacitance in real time to detect the condensation state on the sensor surface;

[0095] Step S22: When the change in resistance or capacitance exceeds a preset threshold, it is determined that condensation has occurred on the surface of the humidity sensor, and a compensation correction procedure is initiated;

[0096] Step S23: Based on a pre-calibrated condensation deviation curve or an adaptive machine learning model, correct the humidity measurement affected by condensation in real time;

[0097] The pre-calibrated condensation deviation curve is obtained by mapping the condensation state to the humidity deviation through experimental calibration, while the adaptive machine learning model is trained on historical condensation data and corresponding real humidity data to predict the humidity deviation based on the condensation monitoring signal.

[0098] Step S24: Use the corrected humidity data to update the input of the time series prediction model, and dynamically adjust the intervention strategies of air pressure regulation, humidity regulation, and static elimination to improve the accuracy of dust deposition anomaly detection.

[0099] In one possible embodiment, in order to compensate for the data deviation of the humidity sensor caused by condensation, a condensation monitoring electrode and a micro-heating element are set on the surface of the humidity sensor to monitor the resistance or capacitance changes in real time to detect condensation; when the monitoring value exceeds a preset value, the micro-heating element is started to heat for a preset time to evaporate the condensation.

[0100] In a possible embodiment, the system includes a pattern matching module, which, based on the dust deposition rate curve generated by the dust deposition risk prediction module and the real-time environmental parameters provided by the environmental data acquisition module, constructs a high-dimensional feature space including humidity-potential-dust multi-dimensional time-lag characteristics and cross-regional airflow characteristics through a time series prediction model and trains the initial control mode. It uses historical operation data and actual deposition data fed back by optical sensors to optimize the generation of a matching relationship library between environmental parameters and operation control modes through supervised learning combined with a clustering algorithm. The matching relationship library records the preferred air pressure adjustment values ​​under different dust concentrations, humidity ranges, and electric field strength conditions. Humidity adjustment value and static elimination strategy; in real-time operation, the pattern matching module queries the corresponding control mode from the matching relationship library according to the current environmental parameters and outputs the intervention strategy to reduce the real-time prediction frequency; if the environmental parameters exceed the coverage of the matching relationship library, the adaptive learning mechanism is triggered, and based on real-time environmental data and intervention effect feedback, the reinforcement learning algorithm is used to dynamically adjust the weight of the time-lag feature in the high-dimensional feature space, and the matching relationship library is updated to incorporate new working conditions; the pattern matching module regularly optimizes the weight parameters of the control mode in the matching relationship library based on the actual deposition data fed back by the optical sensor to ensure that the intervention strategy is consistent with the deposition trend of the dynamic working conditions.

[0101] Summary: Based on Example 1, Example 2 of the present invention introduces a dew point compensation correction program to address the data drift problem of the humidity sensor caused by local condensation. It uses a condensation monitoring electrode to detect the condensation state on the surface of the humidity sensor in real time, and corrects the humidity data in real time based on a pre-calibrated condensation deviation curve or an adaptive machine learning model. It also combines a micro-heating element to evaporate condensation to ensure the accuracy of the humidity data. At the same time, the pattern matching module constructs a high-dimensional feature space containing humidity-potential-dust multi-dimensional time-lag characteristics and cross-regional airflow characteristics, trains a matching relationship library between environmental parameters and operation control modes, and uses reinforcement learning to dynamically adjust feature weights and optimize intervention strategies, which significantly reduces the real-time prediction frequency, improves the accuracy of dust deposition anomaly detection, effectively solves the misjudgment problem caused by humidity sensor drift, and enhances the stability and reliability of the multi-region environmental control system of engineering vehicles in complex dust environments.

[0102] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. AI-based multi-zone environmental control system for engineering vehicles, characterized by: include: Environmental data acquisition module, which collects environmental parameters of each area, including humidity value, electric field strength, air pressure data and dust concentration data, and records the distribution map of dust particle deposition in each area of ​​the engineering vehicle; The dust deposition anomaly identification module extracts the time series characteristics of the dust particle deposition distribution map, calculates the deposition acceleration coefficient and humidity response coefficient, and calculates the dust deposition risk factor based on the deposition acceleration coefficient and humidity response coefficient. It compares the coefficient with the preset value and marks the area exceeding the threshold as an area with abnormal dust deposition status. The dust deposition risk prediction module analyzes the deposition trend of areas experiencing abnormal dust deposition based on a time series prediction model. By inputting environmental parameters, it predicts the dust deposition rate curve for each area of ​​the engineering vehicle within a set timeframe in the future. The input features of the time series prediction model include humidity change rate, potential fluctuation trend, deposition rate change rate, deposition amount trend, cross-regional airflow characteristics, and time lag characteristics. The time series prediction model uses a graph convolution-bidirectional LSTM hybrid network. The graph convolution layer extracts spatial associations based on the airflow adjacency matrix generated by offline CFD simulations, and dynamically adjusts weights online through interpolation to reflect airflow changes. The bidirectional LSTM captures temporal dependencies and predicts the dust deposition velocity curve for each region. When the prediction deviation exceeds the preset value, an adaptive calibration procedure is triggered. The dust deposition intervention module uses a multi-objective optimization algorithm to dynamically calculate a coordinated intervention strategy for air pressure regulation, humidity regulation, and static elimination based on the predicted dust deposition rate curve, effectively reducing the dust deposition rate. The monitoring and adjustment module monitors the condensation status on the surface of the humidity sensor in real time, compensates and corrects the humidity data affected by condensation, and dynamically adjusts the intervention strategy.

2. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: The system includes a regional division module, which divides the engineering vehicle into regions based on the airflow path, equipment layout and dust deposition characteristics within the engineering vehicle; the dust deposition characteristics are obtained by analyzing the air duct path, airflow direction and flow distribution of the airflow path, and K-means clustering is used to partition the engineering vehicle cabin.

3. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: The deposition acceleration coefficient is obtained by calculating the rate of change of dust concentration and air pressure; collecting dust concentration data and air pressure data in each area, and calculating the rate of change of dust concentration, that is, the amount of change in dust concentration per unit time; Calculate the rate of change of air pressure, that is, the amount of change in air pressure per unit time; multiply the rate of change of air pressure by the rate of change of dust concentration to obtain the deposition acceleration coefficient, which represents the acceleration trend of dust deposition; The humidity response coefficient is obtained by calculating the humidity change rate and the average humidity level of the airflow; the humidity change in the area is monitored in real time, the rate of humidity change is calculated, and the average humidity level of the area is calculated; after dimensionless processing, the humidity change rate and the average humidity level of the area are weighted to obtain the humidity response coefficient, which reflects the impact of humidity changes on dust deposition; The dust deposition risk coefficient is obtained by multiplying the deposition acceleration coefficient and the humidity response coefficient.

4. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: The system includes an early warning module connected to the dust deposition risk prediction module, which compares the predicted dust deposition rate curve with the early warning threshold. When the early warning threshold is exceeded, an early warning is issued and the dust deposition intervention module is executed. A response window is defined based on the response time of the intervention strategy and the confidence interval of historical deposition rate data, and the dust deposition amount within the response window is calculated using a time series prediction model. The static basic warning threshold is associated with the rate of change of the predicted deposition rate to generate a dynamic warning threshold function. The warning threshold is tightened within the response window as the dust deposition rate increases. The warning threshold is scaled and corrected according to the real-time humidity difference and electric field strength deviation to suppress the falsely high warning threshold in high humidity and strong electric field environments. When the cumulative deposition reaches a preset proportion or the operating cycle exceeds the limit, the dynamic warning threshold function is recalibrated to eliminate the influence of sensor drift and environmental aging.

5. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: The method for obtaining the time series prediction model includes the following steps: Step S1: Data acquisition and preprocessing: Acquire real-time monitoring time series data of humidity values, electric field strength, dust concentration, current dust deposition rate, and dust deposition amount in each area of ​​the engineering vehicle to form a standardized data sequence to be analyzed; Step S2: Feature extraction: extract the features of the data sequence to be analyzed and output a feature vector, including: humidity change rate; potential fluctuation trend; deposition rate change rate, deposition amount trend, cross-regional airflow characteristics, and time lag characteristics; Step S3: Model construction: Using the feature vector as input, a graph convolution-bidirectional LSTM hybrid network is used. The graph convolution layer extracts spatial associations based on a pre-calculated airflow adjacency matrix. The pre-calculated airflow adjacency matrix is ​​generated through offline CFD simulation and covers the airflow distribution under typical working conditions. In real-time operation, the adjacency matrix weights are dynamically adjusted through online interpolation methods to reflect airflow changes. The bidirectional LSTM captures time series dependencies and adaptively adjusts the adjacency matrix weights in each cycle to reflect turbulent evolution. The dust deposition velocity curves of various areas of the engineering vehicle within a set time in the future are predicted. The prediction error of the dust deposition velocity curve is used as the loss function, and the model parameters are optimized through cross-validation. The output is the dust deposition velocity curve of each area, which shows the trend of the deposition velocity per unit time over time. Step S4: Model verification and calibration: Based on historical operating data and real-time measurement data, the accuracy of the predicted dust deposition velocity curve is verified; when the deviation between the predicted value and the actual measurement value exceeds a preset value, the adaptive calibration procedure is triggered.

6. The AI-based multi-zone environmental control system for engineering vehicles according to claim 4 is characterized in that: The method for obtaining the time-lag feature comprises the following steps: The nonlinear correlations between humidity, potential, deposition velocity, and dust deposition were calculated. The time-lag relationships between each parameter and dust deposition were modeled using mutual information metrics and nonlinear cross-correlation analysis, generating eigenvectors describing the nonlinear time lag between each parameter. A neural network regression model is used, which includes an input layer, one or more hidden layers, and an output layer. The input layer receives humidity, potential, air pressure difference, and deposition rate parameters, the hidden layer uses a fully connected layer, and the output layer predicts the time-lag relationship between humidity, potential, and air pressure difference parameters and dust deposition. Training is performed using a backpropagation algorithm, with the loss function being the mean square error, and optimization is performed using a gradient descent method. Based on real-time monitoring data, the model parameters of the time-lag characteristics are adjusted through an adaptive algorithm. The adaptive algorithm compares the actual monitoring results with the model prediction results at each feedback and adjusts the network weights to ensure the accuracy of the time-lag characteristics. Based on the airflow adjacency matrix generated by CFD simulation, the spatial topological relationship of the airflow in the adjacency matrix is ​​used as input and combined with the fluid dynamics model to optimize the calculation of time-delay characteristics. The accuracy of nonlinear time-delay modeling is improved by optimizing the input characteristics. The calculated time-lag characteristics are fed back to the time series prediction model. When the predicted deposition rate and the actual measurement data exceed the preset error, the adaptive calibration program is triggered to adjust the neural network regression model through a retraining process.

7. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: By deploying micro airflow velocity sensors at the corners of air ducts, narrow spaces or air flow intersections, the velocity gradient and turbulence characteristics of local microairflows can be monitored in real time. After the microairflow disturbance data are fused, the local microairflow disturbance intensity is generated, including the velocity gradient and turbulence characteristics. When the monitoring value exceeds the threshold, the adjustable air valve at the corresponding position or the auxiliary small fan is controlled to adjust the airflow direction or flow rate to weaken turbulence or enhance the scouring effect and reduce dust deposition.

8. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: To address the problem of humidity sensor data drift due to local condensation, which in turn leads to misjudgment of abnormal dust deposition conditions, the dew point temperature is calculated in real time. When the difference between the ambient temperature and the dew point exceeds the preset value, the dew point compensation correction program is triggered to perform real-time compensation correction on the humidity measurement values ​​affected by condensation, thereby eliminating the interference of humidity sensor drift on the prediction results and improving the accuracy of the judgment of abnormal dust deposition conditions.

9. The AI-based multi-zone environmental control system for engineering vehicles according to claim 1 is characterized in that: The multi-objective optimization algorithm includes: Air pressure regulation targets, which adjusts air pressure in real time to optimize airflow and prevent accelerated dust deposition by minimizing the deposition acceleration factor; Humidity adjustment target, by calculating the humidity response coefficient, adjusts the humidity range in real time to prevent excessive humidity fluctuations from causing abnormal dust deposition; Static elimination goal: Based on a comprehensive model of electric field strength and humidity changes, optimize the static elimination strategy to reduce the impact of static electricity accumulation on dust deposition; The air pressure regulation target, humidity regulation target and static elimination target are integrated through a weighted objective function, wherein the weights of the targets are dynamically adjusted according to real-time environmental data and prediction results, thereby achieving coordinated optimization of air pressure, humidity and static elimination.

Images