CN107379899A - A kind of tire condition intelligent monitor system based on wireless sensor network - Google Patents

Abstract

The invention discloses a tire state intelligent monitoring system based on a wireless sensor network. The tire parameter acquisition platform detects tire temperature, wheel acceleration, tire pressure and ambient temperature, and the tire safety state intelligent early warning model outputs the safety state of the tire according to the inspection parameters of the acquisition platform; the invention effectively solves the problem of the existing direct tire pressure detection system. Detect tire safety status parameters such as tire pressure and temperature, but cannot detect other factors that affect tire blowouts such as vehicle operating status, tire quality level, and tire wear.

Description

A tire condition intelligent monitoring system based on wireless sensor network

technical field

The invention relates to the technical field of tire detection equipment, in particular to an intelligent tire state monitoring system based on a wireless sensor network.

Background technique

During the high-speed driving of the vehicle, tire failure is the most worrying and most difficult to prevent for all drivers, and it is also an important reason for sudden traffic accidents. According to the survey and statistics of the American Society of Automotive Industry Engineers, 260,000 traffic accidents in the United States are caused by tire failures every year, and 75% of the tire failures are caused by tire underpressure, overpressure, penetration and excessive tire temperature. Tires are the only parts that are in contact with the ground during the driving process of the car. The tires carry the entire mass of the car, buffer the impact of the road surface, and generate driving force and braking force through the adhesion with the ground. When the car is running with insufficient tire pressure, it will have the following adverse effects: ①Under the same loading conditions, the carcass deformation is large, the tire temperature rises during driving, the rubber is aging, and cord delamination is prone to occur; ②The tire sinkage Large, sunken tires, prone to wear shoulders during use; ③ Large deformation of the tire section, narrowing the distance between twin tires, which is likely to cause sidewall collision wear; ④ Abnormal wear of the tires, reducing tire life, and laying the groundwork for tire blowouts Hidden dangers; ⑤ Increased tire rolling resistance, high fuel consumption, and poor steering performance; during emergency braking, if the tire pressure on one side is low, it will cause the body to deflect and even cause an accident. Tire pressure is an important parameter affecting tire performance. Running a car with insufficient or excessive tire pressure will not only cause the tires to overheat and affect the handling of the car, increase the possibility of a tire blowout, but also reduce the life of the tire and fuel efficiency. The tire pressure monitoring system can automatically warn the driver when one or more tires have abnormal pressure by monitoring the pressure, temperature and wheel motion inside the tire. The indoor test proves that it is generally believed that if the air pressure is increased by 25%, the tire life will be reduced by 15%-20%; if the air pressure is reduced by 25%, the life span will be reduced by about 30%. The higher the temperature of the car tire, the lower the strength of the tire and the greater the deformation (generally the temperature cannot exceed 80°, when the temperature reaches 95°, the condition of the tire is very dangerous), and the tire wear will increase by 2% for every 1° increase; Every time the driving speed doubles, the mileage of the tire will be reduced by 50%. Therefore, overheating and overspeeding are not allowed. The normal tire pressure value of a general car is around 210kPa, and it is advisable for a multi-seat commercial vehicle to be around 240kPa. The early tire pressure detection system is an indirect automobile tire pressure monitoring system, which compares the speed difference between the wheels through the wheel speed sensor of the automobile ABS system, so as to achieve the purpose of indirectly monitoring the tire pressure. The main disadvantage of this type of system is that it cannot judge the situation that two or more tires are insufficient at the same time and the speed exceeds 100km/h. Most of the current tire pressure detection systems are direct automobile tire pressure monitoring systems, which use the pressure sensors installed inside the tires to directly measure the tire pressure, temperature and wheel movement status, and transmit them wirelessly to the receiver installed in the cab. device, it can timely alarm and effectively prevent tire blowout when the tire pressure is too high, too low, slow tire leakage or abnormal temperature changes, the driver can intuitively understand the pressure status of each tire; it can monitor the status of all tires at the same time and Display and monitor the tire pressure. In contrast, the direct tire pressure detection system is superior to the indirect tire pressure detection system both in terms of function and performance, and the current mainstream of the market is the direct tire pressure detection system. In order to improve the coordination of various systems, optimize the overall performance of the car, save costs, and increase the comfort, convenience and safety of the car. But this kind of detection system that only detects tire safety state parameters such as tire pressure and temperature, the driver only knows the single parameter status of the current tire, because the factors that affect the tire blowout involve tire pressure, temperature, car running status, tire quality grade and Tire wear and other factors, this patent designs a tire safety state intelligent monitoring system based on wireless sensor network, realizes intelligent monitoring of tire safety state, and timely forecasts tire safety state parameters to the driver, so as to take relevant measures in time .

Contents of the invention

The present invention provides an intelligent monitoring system for tire status based on a wireless sensor network. Other factors that affect tire blowouts such as quality grades and tire wear.

The present invention is realized through the following technical solutions:

An intelligent tire condition monitoring system based on a wireless sensor network, characterized in that: the intelligent monitoring system includes a tire parameter collection platform based on a wireless sensor network, an intelligent early warning model for tire safety status; the tire parameter collection based on a wireless sensor network The platform detects tire temperature, wheel acceleration, tire pressure and ambient temperature, and the tire safety status intelligent early warning model outputs the tire safety status according to the inspection parameters of the collection platform, among which:

The tire parameter acquisition platform based on the wireless sensor network includes four detection units and a receiving unit for detecting automobile tire parameters. The above-mentioned units are constructed into a wireless sensor measurement and control network through self-organization. The actual values of wheel acceleration and ambient temperature are sent to the receiving unit, and the receiving unit receives the information sent by the detection unit and performs early warning according to the output information of the tire safety state intelligent early warning model;

The tire safety state intelligent early warning model includes the current moment fuzzy least squares support vector machine model, the previous moment fuzzy least squares support vector machine model, GM (1,1) tire temperature prediction model, GM (1,1) Wheel acceleration prediction model, GM(1,1) tire pressure prediction model, next-moment fuzzy least squares support vector machine model, NARX neural network model based on particle swarm optimization and tire safety status classifier, GM(1,1 ) tire temperature prediction model, GM (1,1) wheel acceleration prediction model and GM (1,1) tire pressure prediction model output, tire wear degree and tire quality coefficient are the fuzzy least squares support vector machine model at the next moment Input, the current moment fuzzy least squares support vector machine model, the previous moment fuzzy least squares support vector machine model and the output of the next moment fuzzy least squares support vector machine model are used as the input of the NARX neural network model based on the particle swarm optimization algorithm , the output of the NARX neural network model based on the particle swarm optimization algorithm is the input of the tire safety state classifier, and the tire safety state classifier classifies the safety state of the tire.

The further technical improvement scheme of the present invention is:

The input of the fuzzy least squares support vector machine model at the current moment is the performance parameter of the tire at the current moment, and the fuzzy least squares support vector machine model at the current moment is optimized by the particle swarm optimization algorithm; the fuzzy least squares support vector machine model at the previous moment is The input is the performance parameters of the tire at the previous moment, and the input of the GM(1,1) tire temperature prediction model, GM(1,1) wheel acceleration prediction model and GM(1,1) tire pressure prediction model is the current performance of the tire for a period of time Parameters, the performance parameters of the fuzzy least squares support vector machine model at the previous moment and the fuzzy least squares support vector machine model at the next moment are optimized by the particle swarm optimization algorithm. The tire performance parameters include tire temperature, tire pressure, wheel acceleration, and tire wear and tire quality coefficient.

The further technical improvement scheme of the present invention is:

The output of the current moment fuzzy least squares support vector machine model, the previous moment fuzzy least squares support vector machine model and the next moment fuzzy least squares support vector machine model is used as the input of the NARX neural network model based on particle swarm optimization The NARX neural network model based on the particle swarm optimization algorithm performs continuous dynamic combination prediction on the tire safety performance status at three moments, and the tire safety status classifier classifies the tire safety status according to the output value of the NARX neural network model based on the particle swarm optimization algorithm. These are safe state, comparatively safe state, dangerous state, and critically dangerous state.

Compared with the prior art, the present invention has the following obvious advantages:

One, the present invention is aimed at the fuzzy characteristic of tire safety performance parameter, has set up the fuzzy least squares support vector machine model of current moment, last moment and next moment to detect three tire safety performance states respectively, adopts genetic algorithm to optimize fuzzy minimum A quadratic support vector machine model. The practical application results show that the fuzzy least squares support vector machine model based on the particle swarm optimization algorithm has a relatively small error in predicting tire safety performance and has high prediction accuracy, which provides a powerful theory and technology for fast and effective prediction of tire safety performance analysis support.

Two, the present invention has set up the fuzzy least squares support vector machine model of tire safety performance identification for the fuzziness that affects tire safety performance characteristic parameter, and adopts particle swarm algorithm to the penalty function C of fuzzy least squares support vector machine and the kernel parameter σ are optimized. The relative error of the tire safety performance parameters identified by the fuzzy least squares support vector machine model is less, and the identification accuracy is higher.

Three, the present invention introduces the concept of fuzzy membership degree in the least squares support vector machine, proposes the support vector model based on the fuzzy membership degree function of tire safety performance prediction, gives different degrees of membership according to the degree of input deviation from the data domain to improve the The anti-noise ability of the support vector machine is especially suitable for the situation where the characteristics of the input samples cannot be fully revealed; the experimental and simulation results show that the fuzzy membership function can effectively improve the prediction accuracy of the support vector model of the fuzzy membership function.

Four, the present invention improves the approximate support vector machine by introducing the fuzzy membership degree in order to reduce the impact of singular points and noise on the support vector machine model, and proposes a fuzzy least squares support vector machine model, which retains the support vector machine It has the advantages of strong generalization ability and easy solution, and can overcome the influence of singular points and noise on the approximate support vector machine model. In order to verify the effect of fuzzy support vector machine, this patent conducts empirical research through example data, and compares the experimental results , it can be found that compared with the support vector machine, the fuzzy least squares support vector machine model at the current moment, the previous moment and the next moment has better generalization ability, can significantly reduce the overall error rate, and at the same time improve the Accuracy.

5. The present invention uses the particle swarm optimization algorithm to optimize the parameters of the least squares support vector machine model and the NARX neural network at the current moment, the previous moment and the next moment, and establishes a tire safety performance intelligent prediction model with rapid convergence and high model prediction accuracy. Compared with the traditional traversal optimization method, the modeling efficiency can be greatly improved, and the final model can meet the tire safety performance prediction requirements. This model has certain application value in the analysis and early warning of automobile tire safety performance.

Six, the NARX network model that the present invention adopts is a kind of dynamic recursive network that sets up the NARX network combination model by introducing the time-delay module of tire safety performance characteristic parameter and feedback realization, and it is along the tire safety performance characteristic parameter in time axis direction The extended multi-time tire safety performance characteristic parameter sequence is realized and the data correlation modeling idea of the function simulation function. This method establishes the tire safety performance combination model through the tire safety performance characteristic parameters within a period of time, and the output of the model is The tire safety performance parameters are used as input in the feedback function, and the closed-loop training improves the calculation accuracy of the neural network. The NARX network model realizes the continuous dynamic prediction of the tire safety state.

Seven, the present invention forms the tire safety performance sequence parameter of a time period as the input of the NARX network combination model by setting up the least squares support vector machine model of the tire safety performance prediction of current moment, last moment and next moment respectively, and Compared with a single state, the accuracy and reliability of the NARX network combination model for predicting tire safety performance parameters are improved.

8. The present invention has high prediction accuracy, and combines three GM gray prediction models of tire characteristic parameters pressure, temperature, and wheel acceleration with the least squares support vector machine model and NARX network model of tire safety performance prediction at the next moment to establish a tire The safety performance prediction model makes different choices for the historical data affecting the temperature, pressure, and acceleration of the tire. As the initial data, the GM gray prediction model with 3 parameters is input, and the output of the 3 GM gray prediction models is used as the least squares at the next moment Input to the support vector machine model. The tire safety performance method combines the advantages of less original data and simple method of the GM gray prediction model of gray prediction and the strong nonlinear fitting ability of the least squares support vector machine, and the original data is accumulated and generated through the gray prediction theory , highlighting the impact of the trend, making the least squares support vector machine model to predict the non-linear activation function of tire safety performance easier to approximate, reducing the influence of uncertain components on the predicted value of gray theory; overcoming the disadvantage of low accuracy of gray GM prediction model , which effectively avoids the loss of information in a single model, thereby improving the accuracy of the prediction results; at the same time, the NARX network model is used to predict the state of tire safety performance parameters at three moments, the residual error is small, and the generalization ability of the network is better. The time-to-moment least squares support vector machine model has faster learning time and convergence speed, is more stable, and has higher prediction accuracy. The output of the GM prediction model of tire performance parameters is used as the input of the least squares support vector machine model at the next moment, which improves the accuracy of the output value of the NARX network combination model, thereby greatly improving the accuracy and precision of tire safety performance prediction.

Nine, the present invention has strong robustness, and establishes a gray least squares support vector machine optimized combination automobile tire safety performance prediction model, which embodies the gray system behavior of tire performance parameters, and can dynamically predict, with high precision and stability However, the combination of gray theory, least squares support vector machine and NARX network can make better use of the advantages of each single algorithm, and give full play to the advantages of gray prediction, neural network and least squares support vector machine. Improve prediction accuracy, stability and rapidity; the gray system obtains new data by accumulating or subtracting sample data, which weakens the randomness of the original sample to a certain extent, and has less demand for sample capacity; the patent Combination forecasting can independently learn the inherent laws in the sample data, has strong robustness and fault tolerance, can make more accurate simulation and prediction of tire safety performance, weaken the randomness of original data, and improve the robustness and performance of the prediction model. Fault tolerance is suitable for predicting tire safety performance parameters in various complex situations, and the prediction of tire safety performance parameters has relatively strong robustness.

10. The invention has a long time span for predicting tire safety performance parameters, and the GM gray prediction model can predict the tire temperature, pressure and wheel acceleration in the future according to the temperature, pressure and wheel acceleration that affect the tire safety performance at the previous moment, and input the tire safety state for intelligent early warning The model can predict the tire safety performance parameters in the future. After using the tire S performance parameter values predicted by the above method, the tire temperature, pressure, and wheel acceleration parameter values are added to the original sequence, and a data construction at the beginning of the sequence is correspondingly removed. model to predict tire safety performance parameters. This method is called the equal-dimensional gray number complement model, which can realize long-term forecasting. Users can more accurately grasp the changing trend of tire safety performance, and make full preparations for the safe and reliable operation or maintenance of automobile tires.

11. The present invention improves the scientificity and reliability of tire safety state classification. The tire safety state classifier outputs the NARX neural network model output value according to the degree of influence of tire performance parameters on tire safety performance, expert experience and relevant national standards for automobile tires. Carry out grade division, the different output values of the model correspond to the state of the tire respectively: safe state, relatively safe state, dangerous state and serious dangerous state, realize the classification of tire safety state, and improve the scientificity and reliability of tire safety state.

Description of drawings

Fig. 1 is the tire parameter collection platform based on wireless sensor network of the present invention;

Fig. 2 is a flow chart of the system software of the present invention;

Fig. 3 is the intelligent early warning model of the tire safety state of the present invention;

Fig. 4 is fuzzy least squares support vector machine model of the present invention;

Fig. 5 is a layout diagram of the implementation of the detection unit and the receiving unit of the present invention.

detailed description

1. System hardware design

The tire safety state monitoring system of the present invention is composed of four detection units and a receiving unit. As shown in Figure 1, the detection unit is responsible for accurately measuring the tire parameter information, and the sampling interval is controlled by the single-chip microcomputer and sent to the receiving unit; the receiving unit is responsible for analyzing the signal, running the tire safety state intelligent early warning model, and whether the information output is an alarm. The hardware design of the tire parameter detection unit includes the circuit design of the sensor module, single-chip microcomputer and wireless communication module. The tire parameter detection module adopts the sensor SP12. The sensor SP12 integrates the detection of four parameters such as temperature, pressure, acceleration and ambient temperature. It is composed of a RISC core module with a large number of peripheral devices. The sensor is installed on the root of the valve stem on the rim . It communicates with the AVR ATmega48 MCU through the SPI interface. The SPI interface of the MCU is set to work as the master, and the sensor is set as the slave. The clock signal for synchronous data transmission is provided by the SCK pin of the host MCU, and the wakeup signal of SP12 provides an external interrupt to the MCU. , regularly wake up the MCU to work. The single-chip microcomputer and the wireless transmitter chip TDK5100F realize data communication through the serial port. The PD4 pin of the single-chip microcomputer is connected to ASKDTA. The pin PD5 is connected to PDWN to realize the energy-saving control of the transmitter chip. When PDWN=0, it is the low power consumption mode, and when PDWN=1, it is the working mode. The four units tested adopt the external valve installation method, which is convenient for installation and replacement of tires, and further prolongs the service life of the system. The receiving and displaying part consists of 4 units including AVR ATmega48 single-chip microcomputer, wireless transmitting chip TDK5100F, liquid crystal display and alarm output. It is placed in the driver's cab where it is easy to observe. The alarm adopts two methods of LED light and sound alarm. The intensity of the light and the intensity of the sound are controlled by the size of the alarm level.

2. System software design

This system adopts the method of installing the tire parameter detection unit on the outside of the valve, which is convenient for installation and replacement of tires, prolongs the service life of the system, has high reliability, and is very easy to operate, low in cost, and does not need to add other modules to the system; wireless sensor Installed at the root of the valve stem on the rim, it is mainly used to monitor the internal pressure, temperature and motion status of the tire, and send it to the receiving unit wirelessly. The wireless sensor circuit mainly includes pressure sensors, temperature sensors, acceleration sensors and power supply voltage. The sensor converts the analog signals of tire pressure, temperature, acceleration and voltage into digital signals and sends them to the single-chip microcomputer of the detection unit for data processing and sends them to the receiving unit. The receiving unit processes the received 4 tire parameters and then inputs the tire safety state intelligent early warning algorithm to judge the tire safety state, and displays and warns; the receiving unit sends the tire parameter collection information data frame to the detection unit to continue the detection of the next cycle Unit tire parameter collection and early warning judgment, etc. The system software is programmed in C language. The receiving unit receives the detection parameters of four tires and realizes tire safety status warning judgment and parameter processing. The detection unit realizes tire parameter collection and sends it to the receiving unit. The system working software flow chart is shown in Figure 2 , the tire safety state intelligent early warning model is realized in the software, as shown in Figure 3.

(1), based on fuzzy least squares support vector machine model

According to the actual situation of tire safety performance evaluation, select indicators that are easy to obtain, highly operable, and can most objectively reflect the status quo of tire safety, that is, tire temperature x1, wheel acceleration x2, tire pressure x3, tire wear degree x4, and tire quality coefficient x5 ; Evaluate the safety condition of the tire through the safety factor of the tire to prevent the further deterioration of the safety condition of the tire. In general, the higher the tire temperature x1, the lower the safety factor, the higher the vehicle speed x2, the lower the safety factor, the higher the tire pressure x3, the lower the safety factor, the higher the tire wear x4, the lower the safety factor, and the lower the tire quality factor x5 The lower the safety factor.

①Measurement of fuzzy membership degree: The measurement of fuzzy membership degree u( _xi ,x) is a very important issue, which often directly affects the accuracy of the tire safety performance early warning model of fuzzy least squares support vector machine, membership degree The basis for determining the size is its relative importance in the class. This patent measures the degree of membership based on the distance from the sample to the center of the class. The closer the sample is to the center of the class, the greater the degree of membership, and vice versa, the smaller the degree of membership. The degree function is:

Among them: n _j is the number of sample points belonging to the jth category, and δ>0 prevents the value of the membership function from being zero.

②Current moment fuzzy least squares support vector machine model

In the fuzzy least squares support vector machine, 0<μ(x _k )≤1 represents the fuzzy pre-selection rule after the tire safety state characteristic parameters are fuzzified, which measures the reliability of the sample belonging to a certain category; at the same time, in the least squares In the training process of the multiplication support vector machine, it shows that each training data has a different weight effect on the learning of the least squares support vector machine. Through the fuzzy membership degree, the output value _y1 based on the fuzzy least squares support vector machine model is:

where x=[x ₁ , x ₂ , … x ₅ ], σ is the kernel parameter. The fuzzy least squares support vector machine model is shown in Figure 4.

③ Fuzzy least squares support vector machine model at the previous moment

Based on the fuzzy least squares support vector machine, the fuzzy samples input for tire safety state prediction at the previous moment are tire temperature x1, vehicle speed x2, tire pressure x3, tire wear degree x4 and tire quality coefficient x5. Then the output value of the model based on the fuzzy least squares support vector machine is y ₂ .

④The next moment fuzzy least squares support vector machine model

Based on the fuzzy least squares support vector machine, the fuzzy samples input for tire safety state prediction at the next moment are tire GM gray predicted temperature x1, wheel GM gray pre-acceleration x2, tire GM gray predicted pressure x3, tire wear x4 and tire quality Factor x5. Then the output value of the model based on the fuzzy least squares support vector machine is y ₃ .

The values of y ₁ , y ₂ and y ₃ are used as the input of the NARX neural network combination model.

(2), NARX neural network model design

The NARX neural network is a dynamic recursive neural network with output feedback connections. In terms of topological connection, it can be equivalent to a BP neural network with input delay plus a delay feedback connection from output to input. Its structure consists of input layer, Delay layer, hidden layer and output layer, in which the input layer node is used for signal input, the time delay layer node is used for the time delay of the input signal and output feedback signal, and the hidden layer node uses the activation function to make the delay signal Linear operation, the output layer node is used to linearly weight the output of the hidden layer to obtain the final network output. The output h _i of the i-th hidden layer node of the NARX neural network is:

The output o _j of the jth output layer node of the NARX neural network is:

The input layer, delay layer, hidden layer and output layer of the patented NARX neural network of the present invention are respectively 3-19-10-1 nodes, and the input is respectively the next moment fuzzy least squares support vector machine model, the current moment fuzzy The three outputs of the least squares support vector machine model and the previous moment fuzzy least squares support vector machine model.

(3), particle swarm optimization optimization least squares support vector machine and NARX neural network model

① The general steps of particle swarm optimization least squares support vector machine are as follows:

A. Parameter initialization of particle swarm optimization algorithm. Firstly, determine the penalty parameter and kernel parameter range of the least squares support vector machine, and then determine the relevant parameters of the adaptive particle swarm optimization algorithm. B. Calculate the adaptive weight. C. Calculate and compare the fitness with the minimum sum of regression error squares as the fitness. D. Record the optimal position of each particle and the global optimal position.

② The general steps of particle swarm optimization for NARX neural network combination model are as follows:

A. Parameter initialization of particle swarm optimization algorithm. First, determine the initial weight and threshold of the NARX neural network, and then determine the relevant parameters of the adaptive particle swarm optimization algorithm. B. Calculate the adaptive weight. C. Calculate and compare the fitness with the minimum sum of regression error squares as the fitness. D. Record the optimal position of each particle and the global optimal position.

(4), tire safety status classifier

Tire safety state classifier According to the influence degree of tire performance parameters on tire safety state, expert experience and relevant national standards of automobile tire safety, the output value of NARX neural network model is less than or equal to 0.2, less than or equal to 0.5, less than or equal to 0.7 and less than or equal to 1.0 The values correspond to the status of tires respectively: safe status, comparatively safe status, dangerous status and serious dangerous status, realizing the classification of tire safety status.

3. Embodiment

This system adopts the method of installing the tire parameter detection unit on the outside of the valve, which is convenient for installation and replacement of tires, prolongs the service life of the system, has high reliability, and is very easy to operate, low in cost, and does not need to add other modules to the system; wireless sensor Installed on the root of the valve stem of the valve on the rim, it is mainly used to monitor the internal pressure, temperature and motion status of the tire, and send it to the receiving unit through wireless. The receiving unit is placed where the driver in the cab can easily see it. The plane layout of the receiving unit and the detection unit is shown in Figure 5.

The technical means disclosed in the solutions of the present invention are not limited to the technical means disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features. It should be pointed out that for those skilled in the art, some improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also regarded as the protection scope of the present invention.

Claims (3)

1. A tire condition intelligent monitoring system based on a wireless sensor network, characterized in that: the intelligent monitoring system includes a tire parameter acquisition platform based on a wireless sensor network, a tire safety state intelligent early warning model; the tire condition based on a wireless sensor network The parameter acquisition platform detects tire temperature, wheel acceleration, tire pressure and ambient temperature, and the tire safety status intelligent early warning model outputs the tire safety status according to the inspection parameters of the acquisition platform, among which: The tire parameter collection platform based on the wireless sensor network includes four detection units and a receiving unit for detecting automobile tire parameters, and is constructed into a wireless sensor measurement and control network through self-organization. The detection unit is responsible for detecting automobile tire temperature, tire pressure, wheel The actual values of acceleration and ambient temperature are sent to the receiving unit, and the receiving unit receives the information sent by the detection unit and performs early warning according to the output information of the tire safety state intelligent early warning model; The tire safety state intelligent early warning model includes a fuzzy least squares support vector machine model at the current moment, a fuzzy least squares support vector machine model at the previous moment, a GM (1,1) tire temperature prediction model, and a GM (1,1) Wheel acceleration prediction model, GM (1,1) tire pressure prediction model, next-moment fuzzy least squares support vector machine model, NARX neural network model based on particle swarm optimization and tire safety status classifier, GM (1,1 ) tire temperature prediction model, GM (1,1) wheel acceleration prediction model and GM (1,1) tire pressure prediction model output, tire wear degree and tire quality coefficient are the fuzzy least squares support vector machine model at the next moment Input, the current moment fuzzy least squares support vector machine model, the previous moment fuzzy least squares support vector machine model and the output of the next moment fuzzy least squares support vector machine model are used as the input of the NARX neural network model based on the particle swarm optimization algorithm , the output of the NARX neural network model based on the particle swarm optimization algorithm is the input of the tire safety state classifier, and the tire safety state classifier classifies the safety state of the tire. 2. A kind of tire state intelligent monitoring system based on wireless sensor network according to claim 1, it is characterized in that: the input of described current moment fuzzy least squares support vector machine model is the performance parameter of tire present moment, by particle The group algorithm optimizes the fuzzy least squares support vector machine model at the current moment; the input of the fuzzy least squares support vector machine model at the previous moment is the performance parameters of the tire at the previous moment, GM (1,1) tire temperature prediction model, GM (1 , 1) The input of the wheel acceleration prediction model and the GM (1,1) tire pressure prediction model is the tire’s current performance parameters for a period of time, and the fuzzy least squares support vector machine model at the previous moment and the fuzzy minimum at the next moment are optimized by the particle swarm optimization algorithm. The performance parameters of the quadratic support vector machine model, the tire performance parameters include tire temperature, tire pressure, wheel acceleration, tire wear and tire quality coefficient. 3. A kind of tire state intelligent monitoring system based on wireless sensor network according to claim 1 or 2, it is characterized in that: the fuzzy least squares support vector machine model at the present moment, the fuzzy least squares support vector machine model at the previous moment The output of the machine model and the next-moment fuzzy least squares support vector machine model is used as the input of the NARX neural network model based on the particle swarm optimization algorithm. For dynamic combination prediction, the tire safety state classifier divides the tire safety state into safe state, relatively safe state, dangerous state and serious dangerous state according to the output value of the NARX neural network model based on particle swarm algorithm.