Language translation automobile diagnosis system based on big data

Abstract

The invention discloses a language translation automobile diagnosis system based on big data, and relates to the technical field of automobile diagnosis, the diagnosis system comprises a data acquisition module, a data processing module, a diagnosis analysis module, a language conversion module and a language package module; the data acquisition module comprises a plurality of sensors which are arranged in an automobile and are used for monitoring data in the running process of the automobile in real time; the data processing module preprocesses the collected data, including filtering processing and data conversion processing; and the diagnosis analysis module analyzes the processed data based on a fault diagnosis algorithm and judges whether the automobile has a fault, and the fault diagnosis algorithm learns historical fault data based on a Bayesian network model and determines a probability relationship between each parameter and the fault. The diagnostic analysis module learns massive historical fault data based on a Bayesian network model, and can deeply mine a complex probability relationship between each parameter and a fault, so that the translation quality is continuously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of automotive diagnosis, and specifically to a language translation automotive diagnosis system based on big data. Background Art

[0002] In the current era of rapid development of the automotive industry, the intelligence and safety of automobiles have become the focus of people's attention. With the increasing complexity of automotive electronic systems, accurate and timely diagnosis of automotive faults is crucial for ensuring driving safety and improving automotive performance. Traditional automotive fault diagnosis methods often rely on simple sensor monitoring and experience-based judgments. This approach has limited diagnostic accuracy and is difficult to handle complex fault situations. At the same time, with the integration of the global automotive market, users in different regions have different language requirements for automotive diagnostic reports. However, existing automotive diagnosis systems are usually weak in language translation and cannot meet the diverse language needs of users.

[0003] In the field of language translation, although there are already some translation technologies, for the translation of highly specialized texts such as automotive diagnostic reports, there are still many challenges. On the one hand, there are numerous and constantly updated professional terms in the automotive industry, and ordinary translation systems are difficult to accurately translate these terms. On the other hand, users' requirements for translation quality are increasing day by day, and translation systems need to be continuously optimized based on user feedback. In response to this, we propose a language translation automotive diagnosis system based on big data. Summary of the Invention

[0004] To solve the above technical problems and provide a language translation automotive diagnosis system based on big data, this technical solution solves the above problems.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is: a language translation automotive diagnosis system based on big data, and the diagnosis system includes: a data acquisition module, a data processing module, a diagnostic analysis module, a language conversion module, and a language pack module;

[0006] The data acquisition module includes several sensors, which are arranged inside the vehicle and continuously monitor the data during vehicle operation;

[0007] The data processing module preprocesses the acquired data, including filtering and data conversion;

[0008] The diagnostic analysis module analyzes the processed data based on a fault diagnosis algorithm to determine whether the vehicle has a fault. The fault diagnosis algorithm is based on a Bayesian network model, learns from historical fault data, determines the probability relationship between each parameter and the fault, and gives a diagnostic report;

[0009] The language pack module classifies, indexes, and stores translation data in different languages and updates them in real time;

[0010] The language conversion module includes an optimization unit and a translation unit. The optimization unit regularly collects translation feedback data during the user's usage process, applies big data analysis technology for in-depth mining, and based on the mining results, optimizes the translation algorithm of the online translation interface, adjusts the content and structure of the local language data packet for optimization processing;

[0011] The translation unit translates the given diagnostic report into the language specified by the user.

[0012] Preferably, the positions where the sensors in the data acquisition module are set include the engine, transmission, and braking system. The sensors include temperature sensors and pressure sensors. The temperature sensors are used to collect the engine coolant temperature and oil temperature data in real time, with a collection accuracy of ±1°C; the pressure sensors are used to measure the fuel pressure and intake manifold pressure, with an accuracy of ±0.1 kPa; the sensors continuously collect various real-time data during the operation of the vehicle at a frequency of 100 ms, including but not limited to vehicle speed, engine speed, and exhaust emission components, and transmit the collected data to the data processing module through the CAN bus.

[0013] Preferably, the filtering process is based on the median filtering algorithm. For a set of sampled data with a length of N, it is sorted by size, and the middle value is taken as the filtered data; the data conversion process includes standardization and normalization; the standardization process scales the numerical features to a standard normal distribution with a mean of 0 and a standard deviation of 1; the normalization process is based on the Z-Score method, transforms the original data, and maps it to an interval to complete the normalization process.

[0014] Preferably, after the median filtering algorithm is used for processing, the Kalman filtering algorithm is used to process the data with dynamic changes in vehicle speed and engine speed. By establishing the state space model of the system, the state of the system is predicted and updated. The state equation of the system is:

[0015]

[0016] The observation equation is:

[0017]

[0018] Where X(k) is the system state at time k, A is the state transition matrix, B is the control matrix, U(k) is the control input, W(k) is the process noise, Z(k) is the observation value at time k, H is the observation matrix, and V(k) is the observation noise. The true state of the vehicle system is estimated through the Kalman filtering algorithm.

[0019] Preferably, the diagnostic steps of the fault diagnosis algorithm are specifically as follows: By obtaining historical data, each parameter and fault type in the history are respectively defined as node variables in the Bayesian network. Each node represents a random variable, and its state corresponds to different values of the variable. The dependence relationship between variables is explored in the data, and the cross-validation method is used to verify the constructed network structure; After the structure of the Bayesian network, the conditional probability distribution parameters of each node are estimated according to the sorted historical fault data; For discrete variables, the probability distribution parameters are represented in the form of a table, that is, the conditional probability table, which shows the probability of the node taking each value under different value combinations of the parent node.

[0020] Preferably, during the fault diagnosis process, each parameter value currently monitored by the data acquisition module is input into the constructed Bayesian network model. Using the joint probability distribution formula and Bayes' theorem, under the condition of given observed evidence, the posterior probability of each fault node occurring is calculated, the Bayesian network is calculated and simplified, and the probability values of each fault node in different states are calculated; Based on the expert analysis method, the fault threshold is obtained, and the fault threshold is compared with the probability value for judgment. The fault node exceeding the fault threshold is determined as the fault existing in the current vehicle; A report file of the determined diagnosis is given.

[0021] Preferably, the expression formula sets the fault variable as F, and the parameter variables as C1, C2,..., Cn. According to Bayes' formula:

[0022] P (F|C1, C2,..., Cn)=P(C1, C2,..., Cn|F)*P (F) / P(C1, C2,..., Cn)

[0023] Calculate the probabilities of various faults occurring in the vehicle. When the probability exceeds the set threshold, it is determined that the vehicle has the corresponding fault.

[0024] Preferably, the language pack module classifies various translation data according to the classification rules. After the classification is completed, the language pack module uses the indexing algorithm to establish an index identifier; The vehicle diagnostic system includes a touch display screen, and instructions are input through the touch display screen.

[0025] Preferably, the big data analysis technology in the optimization unit performs clustering analysis on similar translation problems and users to find out the translation errors and user demand patterns, adjusts the translation algorithm of the online translation interface, and performs translation based on the translation preferences of the vehicle system; Among them, clustering analysis divides the data into K clusters to minimize the sum of the squares of the distances from the data points in the cluster to the cluster center. The update formula of the cluster is:

[0026]

[0027] where β r is the updated cluster, qr is the number of data points in the cluster, and g is the data point in q r ; Obtain the translation errors and user requirements in history through cluster analysis.

[0028] Preferably, the translation unit parses the content of the diagnostic report. After the parsing is completed, based on the target language specified by the user, corresponding vocabulary is screened from the language package module, and sentence-by-sentence and paragraph-by-paragraph translation is performed.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0030] The diagnostic analysis module of the present invention learns from a large amount of historical fault data based on the Bayesian network model, and can deeply mine the complex probability relationship between each parameter and the fault. The optimization unit in the language conversion module regularly collects the translation feedback data of users, and uses big data analysis technology for in-depth mining. Based on the mining results, the translation algorithm of the online translation interface is optimized, and at the same time, the content and structure of the local language data packet are adjusted. This continuous optimization mechanism makes the translation quality continuously improved; integrating multiple functional modules such as data acquisition, processing, fault diagnosis, and language translation into one system, realizing a one-stop service from vehicle operation data monitoring to fault diagnosis and then to multi-language presentation of diagnostic reports. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is the framework diagram of the invention diagnostic system. DETAILED DESCRIPTION OF THE INVENTION

[0032] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious variations.

[0033] Refer to Figure 1 As shown, a language translation vehicle diagnostic system based on big data, the diagnostic system includes: a data acquisition module, a data processing module, a diagnostic analysis module, a language conversion module, and a language package module;

[0034] The data acquisition module includes a plurality of sensors, which are arranged inside the vehicle and monitor the data during the operation of the vehicle in real time;

[0035] The data processing module preprocesses the acquired data, including filtering processing and data conversion processing;

[0036] The diagnostic analysis module analyzes the processed data based on a fault diagnosis algorithm to determine whether the vehicle has a fault. The fault diagnosis algorithm is based on the Bayesian network model, learns from historical fault data, determines the probability relationship between each parameter and the fault, and gives a diagnostic report;

[0037] The language pack module classifies, indexes, stores the translation data of different languages, and updates them in real time;

[0038] The language conversion module includes an optimization unit and a translation unit. The optimization unit regularly collects the translation feedback data during the user's usage process, uses big data analysis technology to conduct in-depth mining. Based on the mining results, it optimizes the translation algorithm of the online translation interface, adjusts the content and structure of the local language data packet, and conducts optimization processing;

[0039] The translation unit translates the given diagnostic report into the language specified by the user.

[0040] In the data acquisition module of this application, several sensors installed inside the vehicle can comprehensively and real-time monitor the vehicle operation data. This is like installing "intelligent perception antennas" on the vehicle, covering key parameters such as engine speed, vehicle speed, oil temperature, and oil pressure, presenting the vehicle operation state without reservation. For example, during the vehicle driving process, the sensor can capture the subtle changes in the engine speed in real time, providing accurate first-hand data for subsequent analysis; the data processing module conducts filtering processing and data conversion processing on the collected data. The filtering processing is like "filtering impurities" from the data, removing the noise interference in the signal, making the data purer and more accurate; the data conversion processing can unify the data of different formats and ranges, facilitating subsequent analysis, such as converting analog signals into digital signals, and converting the different data collected by different sensors into a unified standard, improving the data availability;

[0041] Based on the fault diagnosis algorithm of the Bayesian network model, the diagnostic analysis module can accurately determine the probability relationship between each parameter and the fault through learning a large amount of historical fault data. Due to the ability to monitor and diagnose in real time, it can achieve predictive maintenance of the vehicle, give early warnings before the fault occurs, and the vehicle owner can arrange maintenance in time, avoiding the driving safety problems and high maintenance costs caused by sudden faults. For example, if it is predicted that a certain component is about to be damaged, it can be replaced in advance to reduce the impact of the fault on normal use;

[0042] The language pack module classifies, indexes, stores the translation data of different languages, and updates them in real time. This enables the system to meet the needs of users in different regions around the world. No matter what language the user uses, they can obtain the diagnostic report in the corresponding language, greatly improving the universality of the system and the user experience, and helping the product to be promoted in the international market;

[0043] The optimization unit in the language conversion module regularly collects user translation feedback data, deeply mines it using big data analysis technology, optimizes the translation algorithm and adjusts the local language data packet based on the mining results, so as to continuously improve the translation quality. With the development of the automotive industry, new terms emerge continuously. Through user feedback, the translation can be optimized in a timely manner to ensure that the diagnostic report conveys information accurately. For new words emerging in the field of new energy vehicles, the translation can be updated in a timely manner according to user feedback, enabling users to understand without deviation.

[0044] The positions where the sensors in the data acquisition module are set include the engine, transmission, and braking system. The sensors include temperature sensors and pressure sensors. The temperature sensors are used to collect the engine coolant temperature and oil temperature data in real time, with a collection accuracy of ±1°C; the pressure sensors are used to measure the fuel pressure and intake manifold pressure, with an accuracy of ±0.1 kPa; the sensors continuously collect various real-time data during the operation of the vehicle at a frequency of 100 ms, including but not limited to vehicle speed, engine speed, and exhaust emission components, and transmit the collected data to the data processing module through the CAN bus.

[0045] In this application, the sensors are set at the key parts of the vehicle such as the engine, transmission, and braking system, which can directly obtain the operation data of the most important components of the vehicle. The engine is the "heart" of the vehicle, the transmission controls the power transmission, and the braking system is related to driving safety. Monitoring these parts is like installing a "perspective lens" on the core operation links of the vehicle, enabling real-time mastery of its health status. Once the engine shows abnormalities, relevant data changes can be captured immediately, providing a basis for timely maintenance.

[0046] The filtering process is carried out based on the median filtering algorithm. For a set of sampling data with a length of N, it is sorted according to size, and the middle value is taken as the filtered data; the data conversion process includes standardization and normalization; the standardization process scales the numerical features to a standard normal distribution with a mean of 0 and a standard deviation of 1; the normalization process is processed based on the Z-Score method, transforms the original data, and maps it to an interval to complete the normalization process.

[0047] Compared with some linear filtering methods, median filtering can better preserve the edges and sharp features of data while removing noise. Some mutations in vehicle operation data may be important information reflecting critical changes in the operating state, such as the sharp changes in certain parameters at the moment of engine startup. Median filtering will not oversmooth these features like linear filtering, thus ensuring that the important features of the data are not lost, which helps to accurately analyze the instantaneous changes in the vehicle operating state. After mapping the data to a specific interval, the distribution and change trend of the data are more easily visualized and understood. When plotting the graph of vehicle operating parameters changing with time, the normalized data can clearly show the change relationships of each parameter on a unified scale, helping engineers and maintenance personnel to more intuitively analyze the vehicle operating state and quickly discover potential problems.

[0048] After processing with the median filtering algorithm, the Kalman filtering algorithm is used to process the data of the dynamic changes in vehicle speed and engine speed. By establishing the state space model of the system, the state of the system is predicted and updated. The state equation of the system is:

[0049]

[0050] The observation equation is:

[0051]

[0052] Where X(k) is the system state at time k, A is the state transition matrix, B is the control matrix, U(k) is the control input, W(k) is the process noise, Z(k) is the observed value at time k, H is the observation matrix, and V(k) is the observation noise. Through the Kalman filtering algorithm, the true state of the vehicle system is estimated.

[0053] The Kalman filtering algorithm of this application, by establishing the state space model of the system and using the state equation and the observation equation, fully considers the dynamic change characteristics of data such as vehicle speed and engine speed. The state equation describes the state evolution of the system from one moment to the next, while the observation equation correlates the observed value with the system state. This enables the algorithm to more accurately predict and estimate the true state of the system according to the dynamic process of vehicle operation. During the acceleration or deceleration process of the vehicle, the vehicle speed and engine speed do not change in a simple linear manner. The Kalman filtering algorithm can, based on the dynamic model of the system, comprehensively consider the state changes at the previous and subsequent moments and give a more realistic estimated value.

[0054] The diagnostic steps of the fault diagnosis algorithm are specifically as follows: By obtaining historical data, each parameter and fault type in the history are defined as node variables in a Bayesian network. Each node represents a random variable, and its state corresponds to different values of the variable. The dependence relationship between variables is explored in the data, and the cross-validation method is used to verify the constructed network structure. After the structure of the Bayesian network, the conditional probability distribution parameters of each node are estimated according to the sorted historical fault data. For discrete variables, the probability distribution parameters are represented in tabular form, that is, the conditional probability table, which shows the probability of the node taking each value under different value combinations of the parent node.

[0055] In this application, each parameter and fault type in the historical data are defined as node variables in a Bayesian network, comprehensively covering various key information during the operation of the vehicle. This enables the algorithm to deeply mine the laws and relationships hidden behind the historical data. For example, it can discover the potential connections between parameters such as engine temperature and speed and specific fault types. Compared with simple data analysis methods, this approach can make more full use of historical data and provide a rich and valuable information basis for fault diagnosis.

[0056] Using the cross-validation method to verify the constructed network structure effectively improves the reliability of the Bayesian network structure. By dividing and verifying the historical data multiple times, it avoids the construction of incorrect network structures caused by unreasonable data division or model overfitting. This ensures the accuracy and stability of fault diagnosis based on this network structure and reduces the risks of misjudgment and missed judgment.

[0057] During the fault diagnosis process, each parameter value currently monitored by the data acquisition module is input into the constructed Bayesian network model. Using the joint probability distribution formula and Bayes' theorem, the posterior probability of each fault node occurring is calculated under the condition of given observed evidence. The Bayesian network is calculated and simplified to calculate the probability values of each fault node in different states. Based on the expert analysis method, the fault threshold is obtained, and the fault threshold is compared with the probability values for judgment. The fault nodes exceeding the fault threshold are determined as the faults existing in the current vehicle, and a report file of the determined diagnosis is given.

[0058] In this application, the parameter values real-time monitored by the data acquisition module are input into the Bayesian network model, and the posterior probability of the fault node is calculated using the joint probability distribution formula and Bayes' theorem, which can accurately locate the possible faults of the vehicle at present. This process makes full use of the real-time data during the operation of the vehicle and the probability relationships mined from the historical data by the Bayesian network. Compared with traditional empirical judgment or simple rule matching, it greatly improves the accuracy of fault location. For example, in the face of complex fault scenarios where multiple parameters are interrelated and affect each other, the Bayesian network model can comprehensively consider all input parameters and accurately calculate the probability of each potential fault occurring, avoiding misjudgment and missed judgment.

[0059] The expression formula sets the fault variable as F, and the parameter variables as C1, C2, ..., Cn. According to Bayes' formula:

[0060] P (F|C1, C2, ..., Cn)=P(C1, C2, ..., Cn|F)*P (F) / P(C1, C2, ..., Cn)

[0061] Calculate the probabilities of various types of faults occurring in the vehicle. When the probability exceeds the set threshold, it is determined that the vehicle has the corresponding fault.

[0062] The language pack module classifies various types of translation data according to the classification rules. After completion, the language pack module uses an indexing algorithm to establish an index identifier; the vehicle diagnostic system includes a touch display screen, and instructions are input through the touch display screen.

[0063] This application classifies various types of translation data according to the classification rules, which can make the data storage in the language pack module orderly. It classifies according to the diagnostic report terms of different vehicle systems and different language pairs, and groups similar data together. This is like arranging books in a large library according to disciplines, languages, etc., which is convenient for quickly finding the required materials.

[0064] The big data analysis technology in the optimization unit performs clustering analysis on similar translation problems and users to find translation errors and user demand patterns, and adjusts the translation algorithm of the online translation interface for translation based on the translation preferences of the vehicle system; among them, clustering analysis divides the data into K clusters to minimize the sum of the squared distances from the data points within the cluster to the cluster center. The update formula for the cluster is:

[0065]

[0066] where β r is the updated cluster, q r is the number of data points in the cluster, and g is the data point in q r ; Obtain the translation errors and user demand situations in history through clustering analysis; the translation unit parses the content of the diagnostic report. After completion, based on the target language specified by the user, it screens the corresponding vocabulary from the language pack module and performs sentence-by-sentence and paragraph-by-paragraph translation.

[0067] The basic principles, main features and advantages of the present invention have been shown and described above. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, various changes and improvements will occur to the present invention, and all these changes and improvements fall within the scope of the present invention claimed. The scope of protection required by the present invention is defined by the appended claims and their equivalents.

Claims

1. A language translation-based vehicle diagnostic system using big data, characterized in that, The diagnostic system includes: a data acquisition module, a data processing module, a diagnostic analysis module, a language conversion module, and a language pack module; The data acquisition module includes several sensors, which are installed inside the vehicle to monitor the data during vehicle operation in real time; The data processing module preprocesses the acquired data, including filtering and data conversion; The diagnostic analysis module analyzes the processed data based on a fault diagnosis algorithm to determine whether there is a fault in the vehicle. The fault diagnosis algorithm is based on a Bayesian network model, learns from historical fault data, determines the probability relationship between each parameter and the fault, and gives a diagnostic report; The language pack module classifies, indexes, stores, and updates in real time the translation data in different languages; The language conversion module includes an optimization unit and a translation unit. The optimization unit regularly collects the translation feedback data during user use, uses big data analysis technology for in-depth mining, and based on the mining results, optimizes the translation algorithm of the online translation interface, adjusts the content and structure of the local language data packet, and performs optimization processing; The translation unit translates the given diagnostic report into the language specified by the user.

2. The language translation vehicle diagnosis system based on big data according to claim 1, characterized in that: The positions where the sensors in the data acquisition module are installed include the engine, transmission, and braking system. The sensors include temperature sensors and pressure sensors. The temperature sensors are used to collect the engine coolant temperature and oil temperature data in real time, and the acquisition accuracy is ±1°C; the pressure sensors are used to measure the fuel pressure and intake manifold pressure, and the accuracy is ±0.1 kPa; the sensors continuously collect various real-time data during vehicle operation at a frequency of 100 ms, including but not limited to vehicle speed, engine speed, and exhaust emission components, and transmit the collected data to the data processing module through the CAN bus.

3. The language translation vehicle diagnosis system based on big data according to claim 1, wherein: The filtering process is based on the median filtering algorithm. For a set of sampling data with a length of N, it is sorted by size, and the middle value is taken as the filtered data; the data conversion process includes standardization and normalization; the standardization process scales the numerical features to a standard normal distribution with a mean of 0 and a standard deviation of 1; the normalization process is processed based on the Z-Score method, transforms the original data, and maps it to an interval to complete the normalization process.

4. The language translation automotive diagnostic system based on big data according to claim 3, wherein: After the median filtering algorithm is used for processing, the Kalman filtering algorithm is used to process the data with dynamic changes in vehicle speed and engine speed. By establishing the state space model of the system, the state of the system is predicted and updated. The state equation of the system is: The observation equation is: where X(k) is the system state at time k, A is the state transition matrix, B is the control matrix, U(k) is the control input, W(k) is the process noise, Z(k) is the observation value at time k, H is the observation matrix, and V(k) is the observation noise. Through the Kalman filtering algorithm, the true state of the vehicle system is estimated.

5. The language translation vehicle diagnosis system based on big data according to claim 1, characterized in that, The diagnostic steps of the fault diagnosis algorithm are specifically as follows: By obtaining historical data, each parameter and fault type in the history are respectively defined as node variables in the Bayesian network. Each node represents a random variable, and its state corresponds to different values of the variable. The dependence relationship between variables is explored in the data, and the cross-validation method is used to verify the constructed network structure; After the structure of the Bayesian network, the conditional probability distribution parameters of each node are estimated according to the sorted historical fault data; For discrete variables, the probability distribution parameters are represented in tabular form, that is, the conditional probability table, which shows the probability of the node taking each value under different value combinations of the parent node.

6. The language translation automotive diagnostic system based on big data according to claim 5, wherein: During the fault diagnosis process, the parameter values currently monitored by the data acquisition module are input into the constructed Bayesian network model. Using the joint probability distribution formula and Bayes' theorem, under the condition of given observed evidence, the posterior probability of each fault node occurring is calculated, and the Bayesian network is calculated and simplified to calculate the probability values of each fault node in different states; Based on the expert analysis method, the fault threshold is obtained, and the fault threshold is compared with the probability value for judgment. The fault nodes exceeding the fault threshold are determined as the faults existing in the current vehicle; A report file of the determined diagnosis is given.

7. The language translation automotive diagnostic system based on big data according to claim 6, characterized in that The expression formula sets the fault variable as F, and the parameter variables as C1, C2,..., Cn. According to Bayes' formula: P (F|C1, C2,..., Cn)=P(C1, C2,..., Cn|F)*P (F) / P(C1, C2,..., Cn) The probabilities of various faults occurring in the vehicle are calculated. When the probability exceeds the set threshold, it is determined that the vehicle has the corresponding fault.

8. The language translation automotive diagnostic system based on big data according to claim 1, characterized in that: The language pack module classifies various translation data according to the classification rules. After the classification is completed, the language pack module uses the indexing algorithm to establish index identifiers; The vehicle diagnostic system includes a touch display screen, and instructions are input through the touch display screen.

9. The language translation vehicle diagnosis system based on big data according to claim 1, characterized in that, The big data analysis technology in the optimization unit adjusts the translation algorithm of the online translation interface by clustering and analyzing similar translation problems with users to find translation errors and user demand patterns, and performs translation based on the translation preferences of the vehicle system; Among them, clustering analysis divides the data into K clusters to minimize the sum of the squares of the distances from the data points in the cluster to the cluster center. The update formula of the cluster is: where β r is the updated cluster, q r is the number of data points in the cluster, and g is a data point in q r ; historical translation errors and user requirements are obtained through cluster analysis.

10. The language translation vehicle diagnosis system based on big data according to claim 1, characterized in that: The translation unit parses the content of the diagnostic report. After the parsing is completed, based on the target language specified by the user, the corresponding vocabulary is selected from the language pack module and translated sentence by sentence and paragraph by paragraph.

Images