CN110689643A

CN110689643A - Intelligent networking automobile vehicle running state analysis method based on immune algorithm

Info

Publication number: CN110689643A
Application number: CN201910906549.XA
Authority: CN
Inventors: 仝秋红; 刘帅; 吴畏; 杨卓林; 张耀辉
Original assignee: Changan University
Current assignee: Changan University
Priority date: 2019-09-24
Filing date: 2019-09-24
Publication date: 2020-01-14
Anticipated expiration: 2039-09-24
Also published as: CN110689643B

Abstract

The invention discloses an intelligent networking automobile driving state analysis method based on immune algorithm, which comprises the steps of taking an automobile driving state evaluation knowledge base as an autologous base, generating a detector corresponding to each automobile state, taking the autologous base as a known antigen, activating corresponding B cells, generating antibodies corresponding to the antigens by the B cells, activating related antibodies when unknown antigens are generated, and deducing the current state of an automobile according to the probability of activating the antibodies; inputting the distance between the non-self body library and the detector, calculating the distance smaller than a threshold value, determining the probability of each typical state, and evaluating the driving state of the vehicle. The method comprises the steps of acquiring data in the driving process of the intelligent automobile in real time on the basis of an environment sensing system of vision, radar and positioning of the intelligent automobile, establishing a model by using an immune algorithm at a cloud server end, evaluating the driving state of the driving automobile, giving early warning in time when an unsafe state is found, and improving the active prevention and control capacity of road traffic safety.

Description

Intelligent networking automobile vehicle running state analysis method based on immune algorithm

Technical Field

The invention belongs to the technical field of intelligent networked automobile road traffic active safety, and particularly relates to an intelligent networked automobile vehicle running state analysis method based on an immune algorithm.

Background

The automatic driving develops to the present, the monitoring of the driving state of the automatic driving automobile, the safety of the automatic driving automobile is always concerned, the intelligent network connection automobile realizes the environment perception of the driving automobile by the technologies of vehicle-mounted camera, laser radar, millimeter wave radar, satellite positioning and the like, the intelligent algorithm is implemented, some state information parameters of the current automobile in the driving process can be obtained by the vehicle-mounted bus system and other sensors with various functions on the automobile, and the intelligent network connection automobile V2X technology enables the automobile, the people, the automobile, the road, the automobile and the remote monitoring platform to be communicated in real time, so all information on the intelligent automobile can be conveniently transmitted to the related monitoring platform by the technologies, and the driving state of the monitoring automobile can be timely transmitted to the related automobile, People and roads, etc.

Disclosure of Invention

The invention aims to solve the technical problem that the prior art is not enough, and provides an intelligent networking automobile vehicle running state analysis method based on an immune algorithm.

The invention adopts the following technical scheme:

an intelligent networking automobile vehicle running state analysis method based on an immune algorithm comprises the following steps:

s1, taking the automobile driving state evaluation knowledge base as an autologous base, generating a detector corresponding to each automobile state, taking the autologous base as a known antigen, activating corresponding B cells, generating antibodies corresponding to the antigens by the B cells, activating related antibodies when the unknown antigen is generated, and deducing the current state of the automobile according to the probability of activating the antibodies;

and S2, inputting the non-self body library and the detector distance, calculating the distance smaller than a threshold value, determining the probability of each typical state, and evaluating the driving state of the vehicle.

Specifically, step S1 specifically includes:

s101, antigen recognition and self-bank production;

s102, generating B cells, and performing affinity calculation;

s103, updating the memory bank, inhibiting and promoting the antibody, if the affinity is greater than the set threshold, returning to the step S102 to recalculate the affinity, and if the affinity is less than the threshold, generating a detector.

Further, in step S101, the self-set is composed of data of typical vehicle driving states, and as known antigens, a one-dimensional vector set is established from a group of data of vehicle driving states, where the i-th vehicle driving state, j is the number of vehicle state parameters:

code[i]＝{ci1,ci2,ci3……cij}

let c 1: vehicle speed, c 2: acceleration, c 3: battery temperature, c 4: vehicle-to-front vehicle distance, c 5: number of lane departures, which is the number of lane departures a vehicle travels per minute

code[i]＝{ci1,ci2,ci3,ci4,ci5}

N groups of vehicle running state data form an n-dimensional self-body library vector set, i is 1-n, and j is 1-5;

the non-self-body input is any state of the vehicle in the driving process and is used as an unknown antigen:

uncode[i]＝{ui1,ui2……uij}

as with autologous pools:

uncode[i]＝{ui1,ui2,ui3,ui4,ui5}

and n groups of vehicle running state data form an n-dimensional non-self-body library vector set: i is 1 to n, and j is 1 to 5

Further, in step S102, according to the established typical vehicle driving state parameter as an antigen, a certain state parameter is a vector code [ i ] ═ ci1, ci2, … cij ], and a new B cell, Bi (x1, x2 … xj), is generated centering on the antigen; for B cells, a number of N antibodies were generated, assuming that each individual size was k, the number of antibodies was:

N＝j*k

the initial antibody is produced from two sources, if the problem is reserved in the memory bank, the memory bank is taken, the non-foot part is randomly generated, if the memory bank is empty, the whole memory bank is randomly generated, when the antigen invades the body, B cells are activated to recognize the specific antigen, and then the B cells are massively propagated:

B[i]＝{bi1,bi2,…bij}

the new B cells generated by learning are:

newB[i]＝{nbi1,nbi2…nbij}

affinity cloning factors were calculated using Euclidean distance:

further, in step S103, the generated detector corresponding to each vehicle state is:

where i represents the i-th typical vehicle running state, and k is the size of the population.

Specifically, in step S2, after the monitor is generated, with any state of the vehicle running process as a non-self input, the Euclidean distances D between the unknown antigen and the n monitors are calculated, a threshold value M for one distance is set, the cumulative sum M of the distances dij < M in each distance di is calculated, the probability Pi that the antibody belongs to the monitor is obtained, the type of the known antigen to which the unknown antigen belongs is determined from the maximum value of the probability, and the running state of the vehicle is evaluated.

Further, the probability Pi of the monitor is:

Pi＝M/di1+di2+…+dij。

further, defining the assessment to include safe, safer, unsafe and dangerous, the Euclidean distance D of unknown antigen from n monitors is:

D＝{d1,d2…di}

wherein, i is 1 to n, n is 4, di is the distance between the unknown antigen parameter and the ith monitor.

Compared with the prior art, the invention has at least the following beneficial effects:

the invention discloses an intelligent networking automobile driving state analysis method based on an immune algorithm. The immune genetic algorithm adopted by the invention is a multidisciplinary intercross and infiltration optimization algorithm which combines the respective advantages of the immune theory and the basic genetic algorithm, and is applied to the analysis of the driving safety state of the intelligent networked automobile. Finally, the model is verified through experiments, data are analyzed, and the result shows the feasibility of the model for evaluating the driving safety state of the intelligent networked automobile.

Further, an autologous library is established according to the typical vehicle driving state, known antigens are generated, B cells are generated on the basis of the known antigens in the autologous library, an antibody population is generated in a random generation mode, the affinity of the antibody is calculated according to the distance between the antibody and the known antigens, the population is evolved through a genetic algorithm by taking the affinity as the fitness, the population comprises crossing, mutation and selection, an iterative process is controlled through the affinity, and finally a monitor aiming at the typical driving state antigen of each vehicle is generated. The current vehicle running state can be judged by using the generated monitor and the real-time running state parameters of the vehicle as unknown antigens.

Further, taking data of typical vehicle running states as known antigens, establishing a one-dimensional vector set by a group of data of the vehicle running states, wherein the ith vehicle running state, j is the number of vehicle state parameters: code [ i ] = { ci1, ci2, ci3 … … cij }; let c 1: vehicle speed, c 2: acceleration, c 3: battery temperature, c 4: vehicle-to-front vehicle distance, c 5: lane departure times, which are the number of times the vehicle has traveled off the lane per minute: code [ i ] = { ci1, ci2, ci3, ci4, ci5 }; forming an n-dimensional self-body library vector set by n groups of vehicle running state data, wherein i is 1-n, and j is 1-5;

the autologous library is a typical knowledge base for evaluating the driving state of the automobile, and is used as a known antigen of a system to activate B cell learning and produce antibodies corresponding to the antigen.

Further, according to the established typical vehicle driving state parameter as an antigen, a certain state parameter is set as a vector code [ i ] ═ ci1, ci2 and … cij ], and a new B cell, Bi (x1 and x2 … xj), is generated by taking the antigen as a center; for B cells, a number of N antibodies were generated, assuming that each individual size was k, the number of antibodies was: n ═ j × k; when an antigen invades the body, B cells are activated to recognize a specific antigen, and then the B cells multiply in large quantities: b [ i ] - { bi1, bi2, … bij }; the new B cells generated by learning are: newB [ i ] ═ { nbi1, nbi2 … nbij }; affinity cloning factors were calculated using Euclidean distance:

b cells are generated in order to produce antibodies corresponding to antigens, and then the affinity of the antibodies, which represents the principle of matching of the antibodies to the antigens, i.e., the recognition intensity, is calculated as the distance between the antibodies and the known antigens. And (5) leaving the antibody with high affinity for storage to ensure that the population evolves towards a direction with good fitness.

Further, affinity calculated by the distance between the antibody and a known antigen is used as fitness, population is evolved through a genetic algorithm, the population comprises crossing, mutation and selection, an iterative process is controlled through a set threshold value, and finally a monitor aiming at each vehicle typical driving state antigen is generated. The monitor can be used for judging any running state in the running process of the vehicle.

Further, according to the generated monitor, with any state of the vehicle running process as a non-self input, Euclidean distances D between the unknown antigen and n monitors are calculated, a threshold value M of one distance is set, the sum M of the distances of dij < M in each distance di is calculated, the probability Pi of the antibody belonging to the monitor is obtained, the type of the known antigen to which the unknown antigen belongs is judged according to the maximum value of the probability, and the running state of the vehicle is evaluated.

In summary, the autonomously developed intelligent networked automobile vehicle information acquisition system is used as a platform, state parameters of the intelligent automobile in the current running process are acquired in real time and are transmitted to the cloud server, the cloud server establishes a model through an intelligent immune algorithm, the current running state of the intelligent automobile is judged, when an unsafe state is found, early warning is given in time, finally the model is verified through experiments, data is analyzed, and the result shows the feasibility of the model for evaluating the running state of the intelligent networked automobile. The invention is suitable for various traffic conditions and various vehicle types, improves the active safety performance of the automobile in the driving process, and reduces the accident rate. The driver can know the relevant state parameters of the vehicle in real time through the vehicle-mounted terminal, so that the vehicle can be evaluated and maintained conveniently; the manager can also master the vehicle running state through the control center, strengthen the vehicle management and make efficient scheduling.

The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.

Drawings

FIG. 1 is a diagram of an immune evaluation model of a vehicle driving state;

FIG. 2 is a graph of B cell and antibody learning processes;

FIG. 3 is a flow chart of an autologous library generation detector;

FIG. 4 is a flow chart of non-self-body garage input and vehicle driving state evaluation;

FIG. 5 is an input graph of unknown antigens that are not autologous;

FIG. 6 is a graph of population total fitness mean variation;

FIG. 7 is a diagram of the evaluation results.

Detailed Description

Referring to fig. 1 and 2, the biological immune system is a large and complex system, which involves many cellular molecules, and a vehicle driving state immune evaluation model is designed according to the characteristics of the safety of a driving vehicle, and the driving vehicle state is monitored by collecting data on the vehicle and transmitting the data to a cloud server.

The self-body library is a typical automobile driving state evaluation knowledge base and is used as a known antigen of a system, the established typical safety evaluation data is used as the known antigen of the self-body library, corresponding B cells are activated, therefore, antibodies corresponding to the antigens are generated by the B cells, when the unknown antigens are generated, the related antibodies are activated, and the current state of the vehicle is deduced according to the probability of activating the antibodies.

The state space of the intelligent networked automobile vehicle driving state diagnosis immune model is composed of a plurality of main information parameters reflecting vehicle driving, the interaction between an antibody and an antigen can be described, the system state can be represented by a characteristic vector, and corresponding state spaces are respectively established for various states appearing in the solving process.

Referring to fig. 3 and 4, the method for analyzing the driving state of the intelligent networked automobile based on the immune algorithm of the present invention includes the following steps:

s1, generating a detector through the self-body library;

s101, antigen recognition and self-bank production;

initializing autologous library

Inputting the state parameters of a typical vehicle as self-input, initializing a memory bank:

code＝[i][j]

where i is a typical vehicle state and j is a vehicle parameter corresponding to the state.

Because of more states and parameters, several examples are selected, and the evaluation of the vehicle state is divided into 'safe, safer, unsafe and dangerous', the corresponding parameters are selected to 'vehicle speed, acceleration, temperature, distance to the front vehicle and lane departure times', and any state parameter unicode [ i ] of the vehicle running process is used as non-self-input to evaluate the running state.

Self-body set and non-self-body input

The self-set is composed of data of typical vehicle running states, and as known antigens, the data of a group of vehicle running states establish a one-dimensional vector set, wherein the following is the ith vehicle running state, and j is the number of vehicle state parameters.

code[i]＝{ci1,ci2,ci3……cij}

Because the collected vehicle state parameters are many, due to space limitation, the principle of the algorithm is explained by taking several parameters as examples

c 1: vehicle speed, c 2: acceleration, c 3: battery temperature, c 4: vehicle-to-front vehicle distance, c 5: the number of lane departure, which is the number of times the vehicle has traveled off the lane per minute.

According to the example case:

code[i]＝{ci1,ci2,ci3,ci4,ci5}

uncode[i]＝{ui1,ui2……uij}

as with the autologous library, the principles of the algorithm are illustrated by a few of its parameters, then:

uncode[i]＝{ui1,ui2,ui3,ui4,ui5}

S102, generating B cells, and performing affinity calculation;

definition of B cells and Generation of Primary antibodies

According to the previously established typical vehicle driving state parameters as antigens, one of the state parameters is vector code [ i ] ═ ci1, ci2, … cij ], and a new B cell, Bi, (x1, x2 … xj) is generated centering on the antigen.

For B cells, a number of N antibodies were generated, assuming that each individual size was k, the number of antibodies was:

N＝j*k

in the present invention, the number of the exemplified parameters is 5, and the number of the antibodies is 100 if the number of the individual is 20.

The initial antibody is generated from two sources, if the problem is reserved in the memory bank, the memory bank is taken, the random generation is not performed on the part of the antibody, and if the memory bank is empty, the random generation is performed on the whole antibody.

When an antigen invades the body, B cells are activated to recognize a specific antigen, and the B cells proliferate in large numbers.

B[i]＝{bi1,bi2,…bij}

The new B cells generated by learning are:

newB[i]＝{nbi1,nbi2…nbij}

calculation of affinity cloning factors Using Euclidean distance

The affinity represents the matching principle of the antibody and the antigen, namely the recognition intensity, and is calculated according to the distance between the antibody and the center of the B cell, and the general formula for calculating the affinity is as follows:

in the formula, t_kFor the binding strength of antigen and antibody, the binding strength t is calculated by general immune algorithm_kThe math tool of (1) mainly comprises:

haiming distance:

euclidean distance:

the Manhattan distance:

And selecting the antibody with high affinity for storage, and updating the memory bank, so that the antibody with high affinity is obviously promoted, the probability of transmission to the next generation is higher, and the antibody with low affinity is inhibited.

Antibodies

Each B cell is programmed (genetically encoded) to produce a characteristic antibody that is produced by the B cell recognizing an antigen and proliferating and differentiating into plasma cells, and one B cell produces only one specific antibody.

In the learning process, the whole population is removed from low fitness, the remaining population is high in fitness, and the classifier of the ith vehicle state generated after each iteration is

Wherein k is the size of the population, j is the number of the vehicle state parameters, and the number of binary coding bits of each gene is set as c-lemgth.

Genetic manipulation

Because the antibody with high affinity is promoted and the antibody with low affinity is inhibited, the population evolution is easy to be single and local optimization is caused, so a new strategy needs to be introduced into an algorithm to ensure the diversity of the population.

Generation of detectors

The population after cross variation is sorted according to the fitness, the individuals with low fitness are removed, the individuals with high fitness are left, the memory base is updated, and iteration is repeated until the individuals with the maximum fitness in the population.

After the iteration is completed, the generated detector corresponding to each vehicle state is as follows:

S2, inputting the non-self library to evaluate the running state of the vehicle

S201, inputting a non-self distance and a detector distance;

s202, calculating a distance smaller than a threshold value;

and S203, determining the probability of each typical state and finishing the evaluation of the vehicle running state.

Evaluation of any state of the vehicle running process:

when the monitor is generated, the monitor can be used to judge any vehicle running state.

And (3) taking any state of the vehicle driving process as non-self input, calculating the Euclidean distance D between the unknown antigen and n monitors:

D＝{d1,d2…di}i＝1～n

in the present invention, there are 4 evaluations of "safe, safer, unsafe, dangerous", where n is 4 and di is the distance between the unknown antigen parameter and the ith monitor.

Since the number of individuals is k, and j is 1 to k, then:

d1＝{d11,d12,…d1j}

d2＝{d21,d22,…d2j}

：

di＝{di1,di2,….dij}

the invention generates 20 individuals.

Setting a distance threshold M, calculating the cumulative sum M of the distances dij < M in each distance di, thus giving the probability that the antibody belongs to the monitor:

Pi＝M/di1+di2+…+dij

and judging the class of the known antigen to which the unknown antigen belongs according to the maximum value of the probability, thereby evaluating the driving state of the vehicle.

The invention develops an environment sensing system of an intelligent automobile based on vision, radar and positioning, related data in the driving process of the intelligent automobile is obtained in real time through the system, a vehicle-mounted bus and other sensors and is transmitted to a cloud server, a model is established at the cloud server end by an immune algorithm, the driving state of the driving automobile is evaluated, and when an unsafe state is found, early warning is given in time.

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without any inventive step, are within the scope of the present invention.

The invention relates to an intelligent network connection automobile driving state analysis method based on an immune algorithm, wherein experiments are respectively carried out in an intelligent network connection automobile experiment field in the Weishui school district of Changan university and the high speed of the Xian winding city, and experimental equipment and the field are shown in figures 5 and 6.

Experimental facilities include image acquisition and processing system, laser radar range finding and environmental perception system, GPS positioning system, and data source still includes CAN line on the car and other sensors on the car in addition, and all data are all direct or send the high in the clouds server through on-vehicle terminal and handle and save.

The known autologous pools were set as follows:

TABLE 1 autologous library (known antigens)

The 4 groups of data respectively represent that the driving state of the vehicle is safe, safer, unsafe and dangerous.

The data obtained from the experiment were used as input of the unknown antigen, and 50 groups were listed from the experimental data, as shown in fig. 7.

The number k of individual populations is set to be 20, the number c-lemgth of binary coding bits of genes is set to be 4, the initial antibody population is generated randomly, and the average value of the total fitness of the populations is changed as shown in figure 6 after selection, intersection and variation for 100 iterations.

The distance between the population and the self-body library is taken as the fitness, and the fitness is continuously reduced along with the increase of the iteration times, so that the solution is converged.

The probability of each group of data to the known antigen in the self-body library is calculated by the generated monitor for 50 groups of data (the driving state in fig. 7), the state corresponding to the maximum probability is the group of data judgment results, and the obtained judgment results are as shown in fig. 7.

From the figure 10 sets of typical data are listed to illustrate the results of the evaluation, and the unknown antigens are input as shown in table 2:

TABLE 2 input of unknown antigens

The results of the evaluation of the 10 sets of data monitors of table 2 are shown in table 3,

TABLE 3 monitor outputs

The data in the table is the probability that the unknown antigen belongs to the known antigen, and it can be seen from the table that the data 1 is in a safe driving state due to a low vehicle speed and a large distance, the data 6 is in a safer state due to a small distance, and the data 4 has the same probability of two adjacent states, so that the situation is concluded that the situation is positioned between the safe state and the safer state (so that the states are 1.5 and 2.5 in fig. 3), the data 5 is in an unsafe state due to a large acceleration and a large number of deviations, and the data 9 is in a dangerous state due to a high vehicle speed and a large number of deviations.

The situation that the non-adjacent data has the same highest probability also occurs in the output data, the situation is judged to be in a state of low safety, and the data is recorded as data to be further verified.

The invention designs a vehicle running state monitoring system based on an immune algorithm aiming at the characteristics of intelligent networked automobile running states, performs experiments through self-developed equipment to obtain data, takes typical data as a self-body library, takes any running state data as a non-self-body library to evaluate the vehicle running states, verifies that the evaluation of the immune algorithm on the vehicle running states is basically correct, and has great correlation with the integrity of the self-body library, the selection of an initial population, the iteration times and a threshold value.

The above-mentioned contents are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls within the protection scope of the claims of the present invention.

Claims

1. An intelligent networking automobile vehicle running state analysis method based on an immune algorithm is characterized by comprising the following steps:

s1, taking the automobile driving state evaluation knowledge base as an autologous base, generating a detector corresponding to each vehicle state, taking the autologous base as a known antigen, activating corresponding B cells, generating antibodies corresponding to the antigens by the B cells, activating related antibodies when the unknown antigens are generated, and deducing the current state of the vehicle according to the probability of activating the antibodies;

and S2, inputting the distance between the non-self body library and the detector, calculating the distance smaller than a threshold value, determining the probability of each typical state, and evaluating the driving state of the vehicle.

2. The method for analyzing the driving state of the vehicle on the basis of the intelligent networked automobile based on the immune algorithm as claimed in claim 1, wherein the step S1 is specifically as follows:

s101, antigen recognition and self-bank production;

s102, generating B cells, and performing affinity calculation;

3. The method for analyzing vehicle driving states of an intelligent networked vehicle based on an immune algorithm as claimed in claim 2, wherein in step S101, the self-set is composed of data of typical vehicle driving states, as known antigens, a one-dimensional vector set is established from a group of data of vehicle driving states, the ith vehicle driving state, j is the number of vehicle state parameters:

code[i]＝{ci1,ci2,ci3……cij}

code[i]＝{ci1,ci2,ci3,ci4,ci5}

uncode[i]＝{ui1,ui2……uij}

as with autologous pools:

uncode[i]＝{ui1,ui2,ui3,ui4,ui5}

4. The method for analyzing vehicle driving status of an intelligent networked automobile based on immune algorithm as claimed in claim 2, wherein in step S102, according to the established typical vehicle driving status parameters as antigens, a certain status parameter is given as vector code [ i ] ═ ci1, ci2, … cij ], and a new B cell, Bi (x1, x2 … xj), is generated centering on the antigens; for B cells, a number of N antibodies were generated, assuming that each individual size was k, the number of antibodies was:

N＝j*k

the initial antibody is produced from two sources, if the problem is retained in the memory bank, the memory bank is taken, less than part of the antibody is randomly generated, if the memory bank is empty, the antibody is all randomly generated, when the antigen invades the body, B cells are activated to recognize the specific antigen, and the B cells are proliferated:

B[i]＝{bi1,bi2,…bij}

the new B cells generated by learning are:

newB[i]＝{nbi1,nbi2…nbij}

affinity cloning factors were calculated using Euclidean distance:

5. the method for analyzing the driving status of the vehicle on the basis of the intelligent networked automobile based on the immune algorithm as claimed in claim 2, wherein in step S103, the generated detector corresponding to each vehicle status is:

6. The method as claimed in claim 1, wherein in step S2, after the monitor is generated, with any state of the vehicle driving process as a non-self input, the Euclidean distance D between the unknown antigen and n monitors is calculated, a threshold value M of the distance is set, the sum M of the distances dij < M in each distance di is calculated, the probability Pi that the antibody belongs to the monitor is obtained, the class of the known antigen to which the unknown antigen belongs is determined according to the maximum value of the probability, and the driving state of the vehicle is evaluated.

7. The method for analyzing the driving state of an intelligent networked automobile based on an immune algorithm as claimed in claim 6, wherein the probability Pi of the monitor is:

Pi＝M/di1+di2+…+dij。

8. the intelligent networked automobile vehicle driving state analysis method based on immune algorithm as claimed in claim 6, wherein the defined evaluation includes safe, safer, unsafe and dangerous, and Euclidean distance D of unknown antigen from n monitors is:

D＝{d1,d2…di}