An automobile air pump and its dynamic voltage monitoring and control system

Abstract

The present invention relates to the field of automobile air pump monitoring technology, and discloses an automobile air pump and its dynamic voltage monitoring and control system. The system includes a voltage data acquisition module, a dynamic noise suppression module, a voltage fluctuation modeling module, an adaptive compensation control module, etc. The voltage data acquisition module collects and processes multi-dimensional data, the dynamic noise suppression module removes noise, the voltage fluctuation modeling module extracts the voltage and load correlation characteristics, and the adaptive compensation control module optimizes the control parameters to stabilize the voltage. In addition, the system is also equipped with a digital twin simulation module to monitor the air circuit, a distributed update module based on federated learning to optimize the model, a redundant protection circuit to ensure safety, and a fault mode library to provide preventive maintenance suggestions. The system can improve the operating stability, safety and efficiency of automobile air pumps, and effectively solve the shortcomings of traditional air pumps in voltage monitoring and control.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile air pump monitoring, in particular to an automobile air pump and a dynamic voltage monitoring and control system thereof. Background Art

[0002] In the modern automotive industry, the performance and stability of automotive air pumps, as essential equipment for maintaining normal tire pressure, are crucial. With the continuous development of automotive electronic systems, the voltage stability of on-board power supplies faces numerous challenges, significantly impacting the reliable operation of automotive air pumps.

[0003] Traditional automotive air pumps have numerous shortcomings in voltage monitoring and control. First, their onboard power supply voltage monitoring is often inaccurate, preventing real-time and comprehensive voltage data. For example, they only monitor the voltage amplitude while ignoring information such as the frequency and phase of voltage fluctuations, resulting in a delay in detecting potential voltage anomalies. Furthermore, they also have shortcomings in monitoring motor operating current and ambient temperature. Changes in motor operating current are closely related to the load of the air pump. Without accurate monitoring, it is difficult to accurately assess the operating status of the air pump. Ambient temperature also affects the performance of the air pump. For example, low temperatures increase the viscosity of the lubricating oil in the air pump, increasing the load on the motor. However, traditional systems struggle to effectively monitor and address the impact of ambient temperature fluctuations.

[0004] In terms of signal processing, traditional systems lack effective noise suppression. The complex electromagnetic environment inside a vehicle generates significant noise, interfering with the acquisition of signals like voltage and current. This can lead to errors in the acquired data, impacting subsequent analysis and control decisions. Furthermore, the modeling and analysis capabilities for voltage fluctuations are limited, making it impossible to deeply explore the nonlinear correlation between transient voltage fluctuations and load changes. This makes it difficult to accurately predict voltage trends and prevent proactive measures to ensure stable operation of the air pump.

[0005] Traditional automotive air pumps typically use fixed control parameters, unable to dynamically adjust based on real-time data such as voltage, current, and temperature. When the vehicle's power supply voltage fluctuates or the pump load changes, this fixed-parameter control approach can lead to unstable motor speed, affecting inflation efficiency and potentially shortening the motor's lifespan. For example, when the voltage is low, the motor speed may decrease, prolonging inflation time; when the voltage is too high, the motor may overload, causing severe heat and even burning out.

[0006] Furthermore, traditional air pumps also have shortcomings in fault warning and maintenance. Lacking effective fault pattern recognition and prediction mechanisms, repairs are often only possible after a fault occurs, preventing the early detection of potential problems and preventative measures. This inconveniences users, increases operating costs, and poses safety risks. In summary, existing automotive air pump technology suffers from significant deficiencies in voltage monitoring, signal processing, control strategies, and fault warning. A more advanced and intelligent dynamic voltage monitoring and control system is urgently needed to address these issues and improve the overall performance and reliability of automotive air pumps. Summary of the Invention

[0007] The purpose of the present invention is to provide a dynamic voltage monitoring and control system for an automobile air pump to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: a dynamic voltage monitoring and control system for an automobile air pump, the system comprising:

[0009] The voltage data acquisition module is used to collect vehicle power supply voltage, motor operating current and ambient temperature data through multi-channel sensors to generate a multi-dimensional real-time monitoring data set;

[0010] A dynamic noise suppression module is used to perform nonlinear signal decomposition on the multi-dimensional real-time monitoring data set, using an empirical mode decomposition and Kalman filtering algorithm to separate high-frequency noise components and steady-state voltage characteristics, and generate a noise-reduced time series signal sequence;

[0011] A voltage fluctuation modeling module is used to build a deep generative model based on a variational autoencoder, perform latent variable space mapping on the noise-reduced time series signal sequence, extract the nonlinear correlation characteristics between voltage transient fluctuations and load changes, and generate a dynamic voltage state encoding vector;

[0012] The adaptive compensation control module is used to design a strategy optimization network based on reinforcement learning, dynamically adjust the PWM duty cycle and motor speed control parameters according to the dynamic voltage state encoding vector, and generate an anti-disturbance voltage stabilization instruction.

[0013] Preferably, the nonlinear signal decomposition includes:

[0014] Perform empirical mode decomposition on the vehicle power supply voltage signal and generate a set of eigenmode function components through local extreme point interpolation;

[0015] Construct the Kalman filter state equation, use the high-frequency intrinsic mode component as the observation noise input, and iteratively update the system state estimate;

[0016] The sliding window entropy detection algorithm is used to identify the mutation points of the steady-state voltage component and segment the continuous stable signal segments.

[0017] Preferably, the construction of the deep generative model includes:

[0018] A causal convolution layer is introduced into the encoder to capture the temporal causal relationship of the voltage sequence and output the latent variable probability distribution parameters;

[0019] In the decoder, a gated recurrent unit is used to reconstruct the voltage fluctuation waveform, and an attention mechanism is added to focus on the characteristics of the load mutation moment;

[0020] The parameters of the encoder and decoder are optimized through adversarial training strategies to minimize the weighted sum of the reconstruction error and the discriminator loss function.

[0021] Preferably, the design of the strategy optimization network includes:

[0022] The state space is defined as a multidimensional vector of voltage ripple coefficient, current harmonic distortion rate and temperature compensation factor;

[0023] Construct a dual-delay deep deterministic policy gradient algorithm framework, using a target network and experience replay pool to optimize the convergence of the control strategy;

[0024] The reward function is designed as a linear combination of the inverse of the absolute value of the voltage deviation and the energy loss index, driving the policy network to generate the optimal compensation instructions.

[0025] Preferably, the system further comprises:

[0026] The digital twin simulation module is used to establish a motor-air pump coupling model based on physical equations and simulate the dynamic relationship between air chamber pressure and motor torque in real time;

[0027] The measured voltage data is input into the simulation model to generate a virtual pressure curve, and the gas line leakage warning signal is triggered through residual analysis.

[0028] Preferably, the generation of the multi-dimensional real-time monitoring data set includes:

[0029] Synchronously collect power bus CAN signals, Hall sensor current pulses and NTC thermistor temperature data;

[0030] Tensor decomposition algorithm is used to fill missing values ​​in heterogeneous data and construct a three-dimensional spatiotemporal data matrix;

[0031] The sensor baseline drift is eliminated by the moving average difference method, which enhances the identifiability of the transient characteristics of the data.

[0032] Preferably, the system further comprises:

[0033] Build a distributed update module based on federated learning to collaboratively train the voltage prediction model through on-board edge devices and cloud servers;

[0034] A differential privacy mechanism is designed to encrypt local gradient parameters, and a model aggregation algorithm is used to generate a globally shared compensation control strategy.

[0035] Preferably, the generation of the anti-disturbance voltage stabilization instruction includes:

[0036] Establish a multi-objective optimization model with voltage stability, motor life loss and energy efficiency as constraints;

[0037] The non-smooth Newton method is used to solve the optimization problem with inequality constraints and generate the Pareto front solution set of PWM frequency and duty cycle.

[0038] The optimal compromise solution is selected through the fuzzy decision layer and the pulse width modulation signal control parameters are output.

[0039] Preferably, the system further comprises:

[0040] Design a redundant protection circuit based on optocoupler isolation to monitor the junction temperature and switching loss of the IGBT module in real time;

[0041] When an overcurrent or overheating condition is detected, it automatically switches to the backup drive circuit and resets the control strategy parameters.

[0042] Preferably, the system further comprises:

[0043] Build a fault mode library based on time series clustering algorithm and perform feature pattern matching on historical voltage fluctuation data;

[0044] A dynamic time warping algorithm is used to calculate the similarity between real-time signals and fault templates, and to generate preventive maintenance recommendation instructions.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention's dynamic voltage monitoring and control system for automotive air pumps offers significant benefits in multiple areas. In terms of data acquisition and processing, the voltage data acquisition module utilizes multi-channel sensors to synchronously collect vehicle power supply voltage, motor operating current, and ambient temperature data. It utilizes a tensor decomposition algorithm to fill missing values ​​and a moving average difference method to eliminate baseline drift, constructing a precise three-dimensional spatiotemporal data matrix to ensure comprehensive and accurate data input to the system. The dynamic noise suppression module utilizes a combined empirical mode decomposition and Kalman filtering algorithm to effectively separate high-frequency noise components from steady-state voltage characteristics, generating a noise-reduced time series signal sequence. This significantly improves data quality and lays the foundation for subsequent precise analysis.

[0047] The voltage fluctuation modeling module builds a deep generative model based on a variational autoencoder. Through causal convolutional layers, gated recurrent units, and an attention mechanism, it accurately extracts the nonlinear correlation between transient voltage fluctuations and load changes, generating a dynamic voltage state encoding vector. This enables in-depth analysis and effective prediction of voltage fluctuations, providing a strong basis for control strategy formulation. The adaptive compensation control module designs a policy optimization network based on reinforcement learning, dynamically adjusting the PWM duty cycle and motor speed control parameters based on the dynamic voltage state encoding vector. This process optimizes the control strategy by rationally defining the state space, constructing a dual-delay deep deterministic policy gradient algorithm framework, and designing a scientific reward function. This not only ensures voltage stability but also balances motor life loss and energy efficiency. For example, during voltage fluctuations, the PWM duty cycle can be rapidly adjusted to ensure efficient motor operation under stable voltage conditions, reducing additional losses caused by voltage instability, extending motor life, and reducing energy consumption.

[0048] The system's digital twin simulation module establishes a motor-air pump coupling model, simulating the dynamic relationship between air chamber pressure and motor torque in real time. Residual analysis promptly identifies air leaks and issues early warning signals, effectively preventing reduced inflation efficiency and equipment damage caused by leaks and ensuring the proper functioning of the air pump. A distributed update module based on federated learning collaborates with onboard edge devices and cloud servers to train a voltage prediction model. It utilizes differential privacy mechanisms to encrypt local gradient parameters and employs a model aggregation algorithm to generate a globally shared compensation control strategy. This not only protects data privacy but also leverages the cloud's powerful computing power to continuously optimize the model, improving the system's adaptability and performance.

[0049] The redundant protection circuit monitors the junction temperature and switching loss of the IGBT module in real time. When an overcurrent or overheating condition is detected, it automatically switches to the backup drive circuit and resets the control strategy parameters, greatly improving the safety and reliability of the system and preventing the entire air pump system from being paralyzed due to an IGBT module failure. The fault mode library analyzes historical voltage fluctuation data through a time series clustering algorithm, calculates the similarity between the real-time signal and the fault template using a dynamic time warping algorithm, and generates preventive maintenance recommendation instructions to help users detect potential faults in advance, perform maintenance in a timely manner, reduce maintenance costs, and improve the efficiency and safety of the automobile air pump. In short, the present invention comprehensively improves the performance of automobile air pumps from multiple aspects such as data acquisition and processing, voltage monitoring and analysis, control strategy optimization, and fault warning and maintenance, and has extremely high practical value and market prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a working principle diagram of a dynamic voltage monitoring and control system for an automobile air pump according to the present invention;

[0051] Figure 2 Flowchart of nonlinear signal decomposition for dynamic noise suppression module;

[0052] Figure 3 Flowchart for deep generative model building for the voltage fluctuation modeling module;

[0053] Figure 4 Flowchart generated for multi-dimensional real-time monitoring dataset;

[0054] Figure 5 This is a schematic diagram of the structure of the vehicle charging pump;

[0055] Figure 6 for Figure 5 Schematic diagram of the structure after the baffle is omitted;

[0056] Figure 7 This is a schematic diagram of the structure after the main shell is partially removed;

[0057] Figure 8 It is a schematic diagram of the main body shell and tool slot structure. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] See also Figures 1-8 The present invention provides a dynamic voltage monitoring and control system for an automobile air pump, and its overall implementation scheme is as follows:

[0060] The dynamic voltage monitoring and control system of the automobile air pump is mainly composed of a voltage data acquisition module, a dynamic noise suppression module, a voltage fluctuation modeling module and an adaptive compensation control module.

[0061] The voltage data acquisition module uses multi-channel sensors to collect data on the vehicle's power supply voltage, motor operating current, and ambient temperature, generating a multi-dimensional real-time monitoring dataset. These sensors are connected to the vehicle's power supply, motor, and ambient temperature monitoring points, enabling simultaneous data collection. The collected data includes power bus CAN signals, Hall effect sensor current pulses, and NTC thermistor temperature data. This heterogeneous data is processed using a specific algorithm to construct a three-dimensional spatiotemporal data matrix, providing a comprehensive and accurate data foundation for subsequent analysis.

[0062] The dynamic noise suppression module receives the multi-dimensional real-time monitoring data set generated by the voltage data acquisition module and performs nonlinear signal decomposition. Using a combination of empirical mode decomposition and Kalman filtering, the module separates high-frequency noise components from steady-state voltage characteristics, generating a noise-reduced time series signal sequence. This process removes noise interference from the data, highlighting the true characteristics of the voltage signal and providing reliable data for subsequent precise analysis.

[0063] The voltage fluctuation modeling module constructs a deep generative model based on a variational autoencoder, mapping the noise-reduced time series signal into a latent variable space. This operation extracts the nonlinear correlation between transient voltage fluctuations and load changes, generating a dynamic voltage state encoding vector. The encoder and decoder in the deep generative model work together, utilizing techniques such as causal convolutional layers, gated recurrent units, and an attention mechanism to mine complex features in the voltage signal, providing key insights for subsequent control strategies.

[0064] The adaptive compensation control module is designed based on a reinforcement learning-based policy optimization network. Based on the dynamic voltage state encoding vector generated by the voltage fluctuation modeling module, it dynamically adjusts the PWM duty cycle and motor speed control parameters to generate anti-disturbance voltage stabilization commands. By defining an appropriate state space, building an effective algorithmic framework, and designing a reasonable reward function, this module optimizes the control strategy, ensuring stable operation of the automotive air pump under varying voltage and load conditions, ensuring proper operation of the pump and the life and energy efficiency of the motor.

[0065] The implementation of the present invention will be further described below with reference to Examples 1 to 6.

[0066] Example 1:

[0067] In this embodiment, specific implementation details of the voltage data acquisition module and the dynamic noise suppression module are further described.

[0068] The voltage data acquisition module synchronously collects power bus CAN signals, Hall sensor current pulses, and NTC thermistor temperature data. The power bus CAN signal contains various information about the vehicle's power supply, such as voltage amplitude and voltage change trends. These signals are collected and transmitted to the system via the CAN bus interface. The Hall sensor uses the Hall effect to convert the motor's operating current into easily detectable current pulse signals. The frequency and amplitude of these pulse signals are related to the motor's current. By counting and analyzing the pulse signals, the motor's operating current data can be accurately obtained. The NTC thermistor is sensitive to ambient temperature, and its resistance value changes with changes in ambient temperature. By measuring the resistance value of the NTC thermistor and using its characteristic curve, the ambient temperature data can be calculated.

[0069] This heterogeneous data may contain missing values ​​and baseline drift. Tensor decomposition algorithms are used to fill missing values ​​in heterogeneous data. Tensor decomposition algorithms decompose multidimensional tensors into multiple low-dimensional tensor products. By performing tensor decomposition on the collected three-dimensional spatiotemporal data matrix, the inherent structure and correlations of the data can be leveraged to reasonably estimate and fill in missing data. For example, if the voltage data collected at a certain moment is missing, the tensor decomposition algorithm can predict the value of the missing voltage data based on other relevant data before and after that moment and the structural characteristics of the entire data matrix, thereby ensuring data integrity.

[0070] At the same time, sensor baseline drift is eliminated through the moving average differencing method. This method first applies a moving average to the data, calculating the average value of the data within a specific time window. The difference is then subtracted from the current data to obtain the difference data. This effectively eliminates baseline drift caused by various factors and enhances the discernibility of transient data features. For example, if a sensor's baseline slowly rises after prolonged use, the moving average differencing method can highlight the true trend of the data, facilitating subsequent accurate analysis of signals such as voltage and current.

[0071] The nonlinear signal decomposition in the dynamic noise suppression module is performed as follows: First, empirical mode decomposition (EMD) is performed on the vehicle power supply voltage signal. EMD is a decomposition method based on the local characteristic time scale of the signal. It finds local extreme points in the signal and performs cubic spline interpolation on these extreme points to generate a set of eigenmode function components. These eigenmode function components represent the characteristics of the signal at different time scales, including high-frequency noise components and steady-state voltage characteristic components.

[0072] Then, the Kalman filter state equation is constructed, and the high-frequency intrinsic mode component is used as the observation noise input. The Kalman filter is an algorithm that uses the linear system state equation to optimally estimate the system state through the system input and output observation data. In this system, the high-frequency intrinsic mode component is regarded as the observation noise, and the noise is effectively suppressed by continuously iteratively updating the system state estimate. Assume that the system state equation is , the observation equation is ,in yes The system state vector at time , is the state transition matrix, is the control input matrix, is the control input vector, is the process noise vector, is the observation vector, is the observation matrix, is the observation noise vector. In this implementation, the high frequency eigenmode components are mapped to ,Through the iterative calculation of the Kalman filter, the system state estimation value is continuously optimized, thereby separating the steady-state voltage characteristics.

[0073] Finally, a sliding window entropy detection algorithm is used to identify the sudden changes in the steady-state voltage component and segment the continuous, stable signal. This algorithm calculates the entropy of the signal within a specific time window. Entropy reflects the uncertainty and complexity of the signal. A sudden change in the entropy value indicates a possible signal anomaly or state change. By continuously sliding the window and calculating the entropy value, the sudden changes in the steady-state voltage component can be identified. The continuous, stable voltage signal can then be segmented into multiple segments, facilitating more detailed analysis and processing of the voltage signal in each segment.

[0074] Example 2:

[0075] This embodiment focuses on the construction process of the deep generative model in the voltage fluctuation modeling module.

[0076] When building a deep generative model based on a variational autoencoder, a causal convolutional layer is introduced into the encoder to capture the temporal causal relationships of the voltage series. This special convolutional layer performs convolution operations only on the current and previous moments and is not affected by future data. Therefore, it effectively captures the temporal causal relationships of the voltage series. For example, when processing data on voltage variations over time, the causal convolutional layer can predict the current voltage trend based on past voltage values, thereby better understanding the inherent patterns of the voltage series.

[0077] The encoder processes the denoised time series signal sequence through the causal convolution layer and outputs the latent variable probability distribution parameters. Assume that the output latent variable probability distribution parameters are (mean) and (logarithm of variance), these parameters represent the distribution characteristics of the voltage series in the latent variable space. By analyzing these parameters, we can discover some characteristics of the voltage series that are difficult to observe directly from the original data.

[0078] The decoder uses a gated recurrent unit (GRU) to reconstruct the voltage fluctuation waveform. The Gated Recurrent Unit (GRU) is a special type of recurrent neural network unit that effectively addresses the vanishing and exploding gradient problems in traditional recurrent neural networks and better captures long-term dependencies in time series. In the decoder, the GRU gradually reconstructs the voltage fluctuation waveform based on the probability distribution parameters of the latent variables output by the encoder. For example, the GRU can simulate voltage variations at different times based on the characteristic information in the latent variables, thereby restoring a voltage fluctuation waveform that is close to the actual waveform.

[0079] At the same time, the decoder overlays an attention mechanism to focus on features at moments of sudden load changes. This attention mechanism allows the model to automatically prioritize the importance of different components when processing data. In voltage fluctuation modeling, voltage characteristics at the moment of sudden load changes are crucial for analyzing system stability and performance. This attention mechanism allows the model to prioritize features at those moments when reconstructing the voltage fluctuation waveform, enhancing the representation of these key features in the model. For example, when the load suddenly increases or decreases, the attention mechanism increases the weight of the voltage change features at that moment, allowing the model to more accurately capture these changes and reflect them in the reconstructed waveform.

[0080] The parameters of the encoder and decoder are optimized through adversarial training strategies. The adversarial training strategy is to introduce a discriminator, whose function is to distinguish whether the reconstructed voltage fluctuation waveform comes from real data or data generated by the model. The encoder, decoder and discriminator compete with each other. The encoder and decoder try to generate reconstructed waveforms that are closer to the real data to deceive the discriminator; while the discriminator tries to accurately distinguish between real data and generated data. The parameters of the encoder and decoder are optimized by minimizing the weighted sum of the reconstruction error and the discriminator loss function. Assume that the reconstruction error is , the discriminator loss function is , and the weighted coefficients are and , then the optimization objective is During the training process, the parameters of the encoder, decoder, and discriminator are continuously adjusted to make the reconstructed waveform more accurate. At the same time, the model's ability to extract the nonlinear correlation features between transient voltage fluctuations and load changes is improved, ultimately generating an accurate dynamic voltage state encoding vector.

[0081] Example 3:

[0082] This embodiment introduces in detail the design of the strategy optimization network in the adaptive compensation control module and the generation process of the anti-disturbance voltage stabilization instruction.

[0083] In terms of strategy optimization network design, the state space is first defined as a multidimensional vector of voltage ripple coefficient, current harmonic distortion rate, and temperature compensation factor. The voltage ripple coefficient is an important indicator for measuring the degree of voltage fluctuation, reflecting the magnitude of voltage fluctuation within a cycle. The current harmonic distortion rate is used to measure the content of harmonic components in the current. Harmonics can adversely affect the operating performance of the motor, so the current harmonic distortion rate is also a key parameter for evaluating system performance. The temperature compensation factor is introduced to account for the impact of ambient temperature on motor performance. Different ambient temperatures will cause changes in motor parameters such as resistance and inductance, which in turn affects the motor's operating state. The temperature compensation factor can correct for these effects.

[0084] A dual-delayed deep deterministic policy gradient algorithm framework (TD3) is constructed, using a target network and an experience replay pool to optimize the convergence of the control policy. The target network is a replica of the policy optimization network, and its parameter updates are relatively slow. Using the target network can reduce fluctuations during the policy optimization process and improve the stability of the algorithm. The experience replay pool stores the empirical data generated during the policy optimization process, including information such as state, action, reward, and next state. During training, a batch of data is randomly sampled from the experience replay pool for training. This breaks down data correlation, prevents the model from falling into local optimal solutions, and accelerates the convergence of the control policy.

[0085] The reward function is designed to be a linear combination of the inverse of the absolute value of the voltage deviation and the energy loss index. Assume that the voltage deviation is , the energy loss index is , the reward function It can be expressed as ,in Is the weight coefficient, which is used to balance the importance of voltage deviation and energy loss in the reward function. When the voltage deviation is smaller, The larger the value, the higher the reward; at the same time, the lower the energy loss index, The smaller the value of , the higher the reward. Through this reward function design, the driving strategy network generates the optimal compensation command, allowing the system to minimize energy loss while ensuring voltage stability.

[0086] To generate anti-disturbance voltage stabilization commands, a multi-objective optimization model was established, using voltage stability, motor life loss, and energy efficiency as constraints. Voltage stability is critical to ensuring the proper operation of automotive air pumps, motor life loss affects the long-term cost of the pump, and energy efficiency impacts the vehicle's energy consumption. These factors were considered as constraints in the multi-objective optimization model, and their interrelationships were comprehensively considered.

[0087] The non-smooth Newton method is used to solve an optimization problem with inequality constraints, generating a Pareto front solution set for PWM frequency and duty cycle. The non-smooth Newton method is an effective approach for solving non-smooth optimization problems. In this system, the multi-objective optimization problem has some non-smooth constraints and objective functions, making it more efficient to use the non-smooth Newton method. This method generates a series of PWM frequency and duty cycle combinations that satisfy different trade-offs. These combinations constitute the Pareto front solution set. The Pareto front solution set represents the set of all solutions that cannot further optimize a particular objective without sacrificing other objectives.

[0088] Finally, the fuzzy decision layer selects the optimal compromise solution and outputs the pulse width modulation signal control parameters. The fuzzy decision layer evaluates and selects each solution from the Pareto front solution set based on the system's actual needs and operational status. Using fuzzy logic, the fuzzy decision layer comprehensively considers multiple evaluation indicators, such as voltage stability, motor life loss, and energy efficiency. By setting appropriate membership functions and fuzzy rules, each solution is fuzzily evaluated. Ultimately, the optimal compromise solution is selected and the pulse width modulation signal control parameters, such as PWM frequency and duty cycle, are output, achieving stable control of the vehicle air pump.

[0089] Embodiment 4:

[0090] The digital twin simulation module establishes a motor-air pump coupling model based on physical equations, simulating the dynamic relationship between air chamber pressure and motor torque in real time. The motor-air pump coupling model comprehensively considers the electromagnetic characteristics of the motor and the mechanical characteristics of the air pump. From the motor side, the electromagnetic torque equation of the motor is established based on the electromagnetic induction law and the structural parameters of the motor. ,in is the motor electromagnetic torque, is the motor torque constant, is the motor armature current. For the air pump, based on the gas state equation and mechanical principles, the relationship equation between the air chamber pressure and the motor torque is established. Assuming the volume of the air pump is , the initial pressure of the gas is , the temperature is , according to the ideal gas state equation ( is the amount of gaseous substance, is the molar gas constant), combined with the working process of the air pump, considering the piston movement, intake and exhaust process of the air pump, etc., to establish the air chamber pressure The equation changes with time. At the same time, the load torque generated by the air pump on the motor and chamber pressure , the structural parameters of the air pump, etc., and the motor-air pump coupling model is established through these physical equations.

[0091] Measured voltage data is fed into the simulation model to generate a virtual pressure curve. This data, acquired through the voltage data acquisition module, reflects the real-time status of the vehicle's power supply. This voltage data is then fed into the motor-pump coupling model. The model calculates the motor's operating status, such as speed and current, based on the voltage data. Furthermore, based on the operating principle of the air pump, the model calculates the change in air chamber pressure over time, generating a virtual pressure curve.

[0092] The gas circuit leakage warning signal is triggered by residual analysis. Residual analysis is to compare the virtual pressure curve with the actual measured gas chamber pressure curve and calculate the difference between the two, that is, the residual. If there is a leak in the gas circuit, the actual gas chamber pressure will be lower than the pressure value under normal circumstances, causing the residual between the virtual pressure curve and the actual pressure curve to exceed the normal range. Set a residual threshold. When the residual is greater than the threshold, the system determines that there may be a leak in the gas circuit and triggers the gas circuit leakage warning signal. For example, under normal circumstances, the residual between the virtual pressure curve and the actual pressure curve is When the residual exceeds When the system is in operation, it will immediately send out an early warning signal to remind the user to check whether there is a leak in the air circuit so that timely repairs can be made to avoid the air pump's working efficiency being reduced or failing to work properly due to air circuit leakage.

[0093] Example 5:

[0094] This embodiment describes the specific implementation of a distributed update module based on federated learning.

[0095] A distributed update module based on federated learning is constructed, enabling the voltage prediction model to be trained collaboratively between the vehicle's edge devices and cloud servers. The vehicle's edge devices possess sufficient computing and storage capabilities, enabling preliminary local processing of collected data. In this system, the vehicle's edge devices collect data related to the vehicle's air pump, including voltage, motor operating current, and ambient temperature.

[0096] The cloud server possesses powerful computing resources and storage capabilities. In-vehicle edge devices encrypt the feature information of local data and upload it to the cloud server, rather than directly uploading the raw data. This protects user data privacy. The cloud server receives the data features uploaded by multiple in-vehicle edge devices and uses a federated learning algorithm for model training. This algorithm allows all participants to collaboratively train a global model without sharing the original data. During training, the cloud server updates the model parameters uploaded by each in-vehicle edge device and generates a globally shared compensation control strategy using a model aggregation algorithm.

[0097] Design a differential privacy mechanism to encrypt local gradient parameters. Differential privacy mechanism is a technology that protects data privacy during data publishing and analysis. When the on-board edge device trains the voltage prediction model, the calculated local gradient parameters contain sensitive information of the local data. In order to protect these privacy, the local gradient parameters are encrypted using the differential privacy mechanism before uploading them. The differential privacy mechanism adds a certain amount of noise to the gradient parameters, so that even if the attacker obtains the encrypted gradient parameters, it is impossible to accurately infer the information of the original data. Assume that the original gradient parameters are , the added noise is , the encrypted gradient parameter for ,in Random noise is generated based on parameters such as the privacy budget of differential privacy. By properly adjusting the distribution and intensity of the noise, we can maximize data privacy while ensuring model training effectiveness.

[0098] A model aggregation algorithm is used to generate a globally shared compensation control strategy. The model aggregation algorithm is one of the key technologies in federated learning. It integrates the local model parameters trained by each on-board edge device to generate globally shared model parameters. Common model aggregation algorithms include the FedAvg algorithm. In this system, the cloud server receives encrypted local gradient parameters uploaded by multiple on-board edge devices and performs weighted averaging and other operations on these parameters according to the model aggregation algorithm to obtain parameter updates for the global model. A globally shared compensation control strategy is then generated based on the updated global model parameters. This compensation control strategy can be distributed to each on-board edge device, enabling the vehicle air pump to be controlled according to a globally optimized strategy, improving the overall performance and adaptability of the system.

[0099] Example 6:

[0100] In the design of a redundant protection circuit based on optocoupler isolation, optocoupler isolation technology utilizes optical signals transmitted between a light-emitting diode and a photosensor to achieve electrical isolation, effectively preventing electrical interference between different circuits and improving system stability and safety. This redundant protection circuit monitors the junction temperature and switching losses of the IGBT module in real time. As a key component in the power drive circuit of an automotive air pump, the operating status of the IGBT (insulated-gate bipolar transistor) directly impacts the pump's performance and reliability. Excessive junction temperature can degrade IGBT module performance or even damage it, while excessive switching losses not only reduce system efficiency but can also cause overheating.

[0101] To this end, the circuit uses a dedicated temperature sensor and current sensor to obtain the junction temperature data of the IGBT module and the current data reflecting the switching loss. The temperature sensor is closely attached to the heat sink or chip of the IGBT module to measure the junction temperature in real time. The current sensor is connected in series in the main circuit of the IGBT to accurately collect the operating current and then calculate the switching loss through the relevant formula. Using the common switching loss calculation formula For example, represents the switching loss, is the collector-emitter voltage, is the collector current, is the switching frequency, and They are the on-time and off-time respectively. 、 and a known switching frequency And other parameters, the switching loss can be calculated.

[0102] When an overcurrent or overheating condition is detected, the redundant protection circuit will automatically switch to the backup drive circuit. This process is implemented by the control logic circuit, which continuously monitors the data collected by the sensor and compares it with the preset overcurrent threshold and overheating threshold. Once the monitored data exceeds the threshold, the control logic circuit immediately sends a switching signal to switch the main drive circuit to the backup drive circuit through a switching device such as a relay or electronic switch. At the same time, the system will reset the control strategy parameters to adapt to the characteristics of the backup drive circuit to ensure that the air pump can continue to operate stably. For example, the parameters of the backup drive circuit may be slightly different from those of the main drive circuit. Resetting the control strategy parameters can adjust the PWM duty cycle, motor speed control parameters, etc. to ensure that the motor can still work normally under the backup drive.

[0103] When building a fault pattern library based on a time series clustering algorithm, we first collected a large amount of historical voltage fluctuation data. This data covers the voltage variations of automotive air pumps under various operating conditions, including normal operation and varying degrees of faults. The time series clustering algorithm groups voltage fluctuation data with similar trends into the same category, forming different fault mode categories. Common time series clustering algorithms, such as the dynamic time warping (DTW) clustering algorithm, calculate the similarity distance between two time series and group them together.

[0104] For each type of fault mode, its characteristic patterns are extracted and stored in the fault mode library as fault templates. For example, under a certain type of fault mode, the voltage fluctuation may show a periodic downward trend, and the decline amplitude and period have a certain regularity. This regularity is the characteristic pattern of the fault mode. In the actual operation process, the dynamic time warping algorithm is used to calculate the similarity between the real-time signal and the fault template. The dynamic time warping algorithm elastically matches the two time series on the time axis to find the best matching path, thereby calculating the similarity between them. Assume that the real-time voltage fluctuation sequence is , the fault template sequence is , the similarity calculated by the DTW algorithm is , the smaller the similarity value, the more similar the two sequences are.

[0105] When the calculated similarity falls below a preset threshold, the system determines that the real-time signal matches a fault template, indicating a potential fault. The system then generates preventive maintenance recommendations based on pre-defined maintenance recommendations in the fault pattern library. For example, if the system detects real-time voltage fluctuations similar to a template indicating a motor winding short-circuit fault, it will issue a recommendation to inspect and repair the motor windings. This allows action to be taken before the fault worsens, reducing repair costs and increasing the reliability and service life of the vehicle's air pump.

[0106] An automobile air pump includes an air pump body 1 made of high-strength engineering plastic. The interior of the air pump body 1 adopts a modular design and is provided with an air pump assembly 2 and a detachable lithium battery pack 3.

[0107] The air pump body 1 also integrates a tire pressure detection module connected to the air pump assembly 2 through a high-pressure air pipe. The tire pressure detection module uses a digital pressure sensor with a detection accuracy of ±0.1PSI.

[0108] The tire pressure detection module is electrically connected to a high-brightness LED display module 4 via a PCB circuit board. An opening 30 is provided on the outer surface of the air pump body 1 for the display surface of the display module 4 to be exposed, and a waterproof rubber ring is provided at the edge of the opening 30.

[0109] Furthermore, the exterior of the air pump body 1 is ergonomically designed, with a slot 10 for placing tool components and the air nozzle. The slot 10 is located on the side of the housing for easy operation. The slot 10 is an embedded structure with a depth of 15-20mm. The slot 10 is also provided with a plurality of tool grooves for placing tool components and a groove for placing the air nozzle. The tool grooves include a dedicated wrench groove, a tire repair tool groove, and a fuse groove. The shape of each groove matches the corresponding tool.

[0110] A detachable baffle 5 is provided outside the notch 10 , and the baffle 5 is fixed by magnetic attraction. A shockproof sponge layer is provided inside the baffle 5 .

[0111] Preferably, the air pump body 1 is provided with a non-slip rubber pad at the bottom and a portable handle at the top; the battery pack 3 supports Type-C fast charging and is provided with a power indicator light; the display module 4 can switch to display real-time tire pressure, preset tire pressure and battery power information.

[0112] Its working principle can be divided into the following parts:

[0113] 1. Air pump inflation system

[0114] Air pump component 2: The motor drives the piston or diaphragm structure to compress the air, generating a high-pressure airflow with a maximum air pressure of 150PSI, which is suitable for inflating car tires, bicycles, balls, etc.

[0115] Inflation nozzle: connected to the air pump through a high-pressure air pipe and adopts a standard valve nozzle interface to ensure close connection with the tire valve to prevent air leakage.

[0116] 2. Tire pressure detection and display system

[0117] Tire pressure detection module: Built-in digital pressure sensor (accuracy ±0.1PSI) monitors tire pressure in real time and transmits data to the control circuit.

[0118] Display module 4: Displays real-time tire pressure, preset tire pressure and battery power through a high-brightness LED screen. Users can adjust the inflation target value according to their needs to achieve precise inflation.

[0119] 3. Power supply system

[0120] Removable battery pack 3: Provides power, supports Type-C fast charging, and has a power indicator to help users understand the remaining power.

[0121] 4. Accessibility Design

[0122] Tool storage slot 10: embedded structure, containing grooves for special wrenches, tire repair tools, fuses, etc., for convenient emergency use.

[0123] Shockproof baffle 5: It adopts a magnetic detachable design and has a shockproof sponge inside to protect tools and reduce vibration noise.

[0124] Workflow:

[0125] Start detection: Connect the tire valve, and the tire pressure detection module automatically reads the current tire pressure and displays it.

[0126] Set target value: The user sets the desired tire pressure (such as 36PSI) through the display module.

[0127] Automatic inflation: The air pump starts and continues to inflate until the set value is reached, then stops automatically to avoid overfilling.

[0128] Storage management: After use, tools and gas nozzles can be returned to the slots and the baffle closed to keep them tidy.

[0129] In summary, the present application uses a built-in tire pressure detection module, and the air pump can not only inflate the tire, but also monitor the tire pressure to ensure that the tire is within the appropriate pressure range, thereby improving driving safety; and the display module 4 is exposed through the opening 30, and the user can directly view the tire pressure data outside the air pump, which is easy and intuitive to operate; at the same time, the slot 10 provides a convenient storage solution, so that tool parts and inflation nozzles have dedicated storage space, keeping the equipment tidy, making the air pump more practical, and allowing users to quickly access required tools and nozzles.

[0130] At the same time, the battery pack 3 and the air pump assembly 2 are electrically connected through the electrical interface terminal 12, which can provide the necessary power for the air pump assembly; the air pump assembly 2 and the tire pressure detection module are connected through an air path, so that the air pump assembly 2 can provide compressed air to the tire pressure detection module to monitor and adjust the tire pressure; the tire pressure detection module and the display module 4 are electrically connected, and the tire pressure detection module can transmit the monitored tire pressure information to the display module 4 for the user to view.

[0131] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0132] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A dynamic voltage monitoring and control system for an automobile air pump, characterized in that: The system comprises: The voltage data acquisition module is used to collect vehicle power supply voltage, motor operating current and ambient temperature data through multi-channel sensors to generate a multi-dimensional real-time monitoring data set; a dynamic noise suppression module for performing nonlinear signal decomposition on the multi-dimensional real-time monitoring data set, using a combined empirical mode decomposition and Kalman filtering algorithm to separate high-frequency noise components from steady-state voltage characteristics, and generating a noise-reduced time series signal sequence including the vehicle power supply voltage, motor operating current, and ambient temperature; The nonlinear signal decomposition comprises: Perform empirical mode decomposition on the vehicle power supply voltage signal and generate a set of eigenmode function components through local extreme point interpolation; Construct the Kalman filter state equation, use the high-frequency intrinsic mode component as the observation noise input, and iteratively update the system state estimate; The sliding window entropy detection algorithm is used to identify the mutation points of the steady-state voltage component and segment the continuous stable signal segments; A voltage fluctuation modeling module is used to build a deep generative model based on a variational autoencoder, perform latent variable space mapping on the noise-reduced time series signal sequence, extract the nonlinear correlation characteristics between voltage transient fluctuations and load changes, and generate a dynamic voltage state encoding vector; The construction of the deep generative model includes: A causal convolution layer is introduced into the encoder to capture the temporal causal relationship of the voltage sequence and output the latent variable probability distribution parameters; In the decoder, a gated recurrent unit is used to reconstruct the voltage fluctuation waveform, and an attention mechanism is added to focus on the characteristics of the load mutation moment; Optimize the parameters of the encoder and decoder through adversarial training strategy to minimize the weighted sum of the reconstruction error and the discriminator loss function; The adaptive compensation control module is used to design a strategy optimization network based on reinforcement learning, dynamically adjust the PWM duty cycle and motor speed control parameters according to the dynamic voltage state encoding vector, and generate an anti-disturbance voltage stabilization instruction.

2. The automobile air pump dynamic voltage monitoring and control system according to claim 1, characterized in that: The strategy optimizes the design of the network, including: The state space is defined as a multidimensional vector of voltage ripple coefficient, current harmonic distortion rate and temperature compensation factor; Construct a dual-delay deep deterministic policy gradient algorithm framework, using a target network and experience replay pool to optimize the convergence of the control strategy; The reward function is designed as a linear combination of the inverse of the absolute value of the voltage deviation and the energy loss index, driving the policy network to generate the optimal compensation instructions.

3. The automobile air pump dynamic voltage monitoring and control system according to claim 1, characterized in that: The system further comprises: The digital twin simulation module is used to establish a motor-air pump coupling model based on physical equations and simulate the dynamic relationship between air chamber pressure and motor torque in real time; The measured voltage data is input into the simulation model to generate a virtual pressure curve, and the gas line leakage warning signal is triggered through residual analysis.

4. The automobile air pump dynamic voltage monitoring and control system according to claim 1, characterized in that: The generation of the multi-dimensional real-time monitoring data set includes: Synchronously collect power bus CAN signals, Hall sensor current pulses and NTC thermistor temperature data; Tensor decomposition algorithm is used to fill missing values ​​in heterogeneous data and construct a three-dimensional spatiotemporal data matrix; The sensor baseline drift is eliminated by the moving average difference method, which enhances the identifiability of the transient characteristics of the data.

5. The automobile air pump dynamic voltage monitoring and control system according to claim 1, characterized in that: The system further comprises: Build a distributed update module based on federated learning to collaboratively train the voltage prediction model through on-board edge devices and cloud servers; A differential privacy mechanism is designed to encrypt local gradient parameters, and a model aggregation algorithm is used to generate a globally shared compensation control strategy.

6. The automobile air pump dynamic voltage monitoring and control system according to claim 1, characterized in that: The generation of the anti-disturbance voltage stabilization instruction includes: Establish a multi-objective optimization model with voltage stability, motor life loss and energy efficiency as constraints; The non-smooth Newton method is used to solve the optimization problem with inequality constraints and generate the Pareto front solution set of PWM frequency and duty cycle. The optimal compromise solution is selected through the fuzzy decision layer and the pulse width modulation signal control parameters are output; The system further comprises: Design a redundant protection circuit based on optocoupler isolation to monitor the junction temperature and switching loss of the IGBT module in real time; When overcurrent or overheating is detected, it automatically switches to the backup drive circuit and resets the control strategy parameters; The system further comprises: Build a fault mode library based on time series clustering algorithm and perform feature pattern matching on historical voltage fluctuation data; A dynamic time warping algorithm is used to calculate the similarity between real-time signals and fault templates, and to generate preventive maintenance recommendation instructions.

7. An automobile air pump, applied to the automobile air pump dynamic voltage monitoring and control system according to any one of claims 1 to 6, comprising an air pump body (1), characterized in that: An air pump assembly (2) and a battery pack (3) are provided inside the air pump body (1). A tire pressure detection module connected to the air path of the air pump assembly (2) is also provided inside the air pump body (1). The tire pressure detection module is electrically connected to a display module (4). An opening (30) is provided on the outer surface of the air pump body (1) for placing a display surface of the display module (4). A notch (10) is also provided on the outside of the air pump body (1) for placing a tool component and an air inflating nozzle therein.

8. The automobile air pump according to claim 7, characterized in that: The notch (10) is an embedded structure, and a plurality of tool grooves for placing tool parts therein and a groove for placing an air inflating nozzle therein are respectively provided in the notch (10), and a detachable baffle (5) is provided on the outside of the notch (10).

Images