Automobile software anomaly deep learning detection method based on service grid architecture

By deploying an end-to-end deep learning anomaly detection model in the service mesh architecture, combining multi-knowledge fusion and adaptive algorithms, the problem that traditional methods are difficult to cope with high-dimensional, strong correlation and nonlinear behavior patterns of vehicle-mounted software is achieved. Effective detection and continuous learning of vehicle-mounted software abnormalities improve detection accuracy and adaptability.

CN119088670AActive Publication Date: 2024-12-06CHINA AUTOMOTIVE CHIP (SHENZHEN) TECHNOLOGY CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202410834702.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-26
Publication Date
2024-12-06
Estimated Expiration
2044-06-26

AI Technical Summary

Technical Problem

Traditional anomaly detection methods based on static rules or simple statistical models are difficult to effectively deal with the high-dimensional, strong correlation, and nonlinear behavior patterns of in-vehicle software, especially in the context of increasing complexity of automotive software systems and increasing complexity, and abnormal states. The pattern is difficult to define in advance.

Method used

Adopting a deep learning anomaly detection method based on the service mesh architecture, by deploying an end-to-end deep learning anomaly detection model on the automotive service mesh control surface, subscribing and analyzing service runtime telemetry data in real time, combining multi-knowledge fusion and adaptive algorithms, Learn normal and abnormal modes to achieve the detection mode upgrade from single structured data to global multimodal data.

Benefits of technology

It realizes effective detection of vehicle-mounted software abnormalities, alleviates the problems of scarcity of abnormal labels and insufficient generalization of models, improves the accuracy and adaptability of software abnormal detection, and can continuously learn and improve, and adapt to complex vehicle-mounted software environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119088670A_ABST
    Figure CN119088670A_ABST
Patent Text Reader

Abstract

The invention discloses an automobile software anomaly deep learning detection method based on a service grid architecture, and the method comprises the following steps: 1, building a lightweight automobile service grid covering an automobile electronic and electrical architecture, and forming a data surface and a control surface which are controlled in a centralized manner; and 2, deploying an automobile software anomaly detection module in an automobile service grid control plane, subscribing and analyzing telemetry data during service operation in real time, and carrying out real-time anomaly detection. The invention provides an automobile software anomaly deep learning detection method based on a service grid framework, which fully utilizes global data convergence capability provided by a control plane in the service grid framework, and isomerous data, such as vehicle state data, environmental parameters and the like, dispersed in each service node is subjected to data processing to obtain a deep learning result. And the data are concentrated to a unified data plane for correlation analysis, so that the detection mode upgrade from single structured data to global multi-modal data is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle-mounted software reliability assurance, and in particular to a software anomaly detection method based on a deep learning algorithm and deployed on an automobile service grid architecture, and in particular to an automobile software anomaly deep learning detection method based on a service grid architecture. Background Art

[0002] As cars transform into software-defined vehicles (SDVs), traditional distributed automotive electronic and electrical architectures have gradually exposed problems such as insufficient scalability, integration difficulties, and data silos, making it difficult to support the complex software functions required by smart cars. To meet this challenge, the next generation of automotive electronic and electrical architectures is evolving toward a centralized, high-performance computing platform. Under the new architecture, electronic control units scattered throughout the vehicle will be gradually integrated into fewer but more powerful domain controllers or even central computing platforms. At the same time, the internal network of the car will also be upgraded to a high-speed interconnection solution based on Ethernet.

[0003] With the transformation of automotive electronic and electrical architecture, the development paradigm of in-vehicle software is also changing. More and more automakers are beginning to introduce advanced IT concepts, decouple software development from hardware, and adopt a microservice architecture model. A smart car software is divided into many loosely coupled microservices, which are developed, tested, and deployed separately, and then dynamically orchestrated on a service grid infrastructure to cope with the explosive growth of software scale and the need for rapid iteration. As a dedicated infrastructure layer, the service grid is responsible for service discovery, communication routing, monitoring and tracking, etc., allowing the microservice development team to focus on business logic and improve development efficiency.

[0004] However, the growing scale and complexity of automotive software systems, as well as the dynamic characteristics introduced by service-oriented architectures, have brought new challenges to software reliability assurance, among which anomaly detection is particularly critical. Traditional anomaly detection methods based on static rules or simple statistical models are difficult to effectively deal with the high-dimensional, strongly correlated, and nonlinear behavior patterns of current in-vehicle software. Abnormal symptoms are often manifested in the interaction data of multiple services, with complex spatiotemporal dependencies and semantic associations, which are difficult to exhaust with a single dimension. In addition, the automobile usage environment is complex and changeable, and the behavior of in-vehicle software is also constantly evolving, making it difficult to define the pattern of abnormal states in advance. Summary of the invention

[0005] In view of the technical problems pointed out in the background technology, the purpose of the present invention is to provide a deep learning detection method for automobile software anomalies based on a service grid architecture.

[0006] To achieve the purpose of the present invention, the technical solution provided by the present invention is specifically as follows:

[0007] A deep learning detection method for automobile software anomalies based on a service grid architecture comprises the following steps:

[0008] Step 1: Build a lightweight automotive service grid covering the automotive electronic and electrical architecture to form a centrally managed data and control plane;

[0009] Step 2: Deploy the automotive software anomaly detection module in the automotive service mesh control plane to subscribe to and analyze service runtime telemetry data in real time for real-time anomaly detection.

[0010] The service runtime telemetry data includes: service call topology data, service input and output data, and resource usage data.

[0011] Among them, the automobile software anomaly detection module includes an end-to-end deep learning anomaly detection model, and uses the anomaly detection model to perform real-time anomaly detection.

[0012] Compared with the prior art, the present invention has the following beneficial effects:

[0013] The present invention proposes a deep learning detection method for automobile software anomalies based on a service mesh architecture. The method makes full use of the global data aggregation capability provided by the control plane in the service mesh architecture, centralizes the heterogeneous data scattered in various service nodes, such as vehicle status data, environmental parameters, etc., to a unified data plane for correlation analysis, and realizes the upgrade of the detection mode from single structured data to global multimodal data.

[0014] In the data-based end-to-end deep learning model, the automotive software domain knowledge is introduced for integration, which alleviates the problems of scarce anomaly labels and insufficient model generalization; and realizes the development of the automotive software anomaly detection field from rule-based to data-driven and then to knowledge fusion. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic diagram of a method flow chart provided in an embodiment of the present application;

[0016] Figure 2 A schematic diagram of an end-to-end deep learning anomaly detection model provided in an embodiment of the present application; DETAILED DESCRIPTION

[0017] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0018] The solution of the present invention fully utilizes the data advantages of the service grid architecture, captures and integrates the multi-dimensional and dynamically migrating behavioral characteristics of vehicle-mounted software operation from a global perspective, and uses adaptive algorithms to learn normal and abnormal patterns. The deep anomaly detection model based on multi-knowledge fusion proposed by the present invention on the service grid automotive electronic and electrical architecture can effectively cope with the above challenges.

[0019] On the one hand, by using the global perspective and data aggregation capabilities of the vehicle software operation provided by the service mesh control plane, key data such as vehicle status parameters and performance parameters of the vehicle software service are natively concentrated on the control plane, forming high-dimensional data features for abnormal analysis. This creates conditions for data-driven intelligent algorithms such as deep learning.

[0020] On the other hand, the end-to-end deep neural network with multi-knowledge fusion can fully explore the complex associations and unknown patterns in the heterogeneous data of the service grid. At the bottom of the algorithm, graph neural networks can be used to extract abnormal signals propagated in the service interaction topology, and time series models can be used to extract dynamic abnormal patterns of single indicators. At the algorithm application layer, the importance weights of different dimensional features for abnormality discrimination are adaptively learned. At the same time, typical abnormal scenario data and physical models based on vehicle software domain knowledge can also be embedded in the network structure to guide the training and reasoning of the anomaly detection model. This end-to-end model can continuously learn from the massive operating data of the service grid and approach the normal and abnormal discrimination boundaries of vehicle software behavior in real environments.

[0021] like Figure 1 As shown, this embodiment provides a method for detecting abnormalities in automobile software by deep learning based on a service grid architecture, comprising the following steps:

[0022] Step 1: Build a lightweight automotive service grid covering the automotive electronic and electrical architecture to form a centrally managed data and control plane;

[0023] It should be noted that the lightweight service grid covering the automotive electronic and electrical architecture is built to form a centrally managed data plane and control plane. Services access the grid through the data plane proxy, and the control plane is responsible for service registration discovery, routing strategy, telemetry monitoring, etc.

[0024] The specific implementation process includes:

[0025] (1) Optimize the service registry to support dynamic registration and discovery of in-vehicle services. Consider introducing a distributed consensus protocol (Raft) to improve the fault tolerance of the service registry.

[0026] (2) Customized service gateway to provide security authentication and protection for off-vehicle access, and fine-grained permission control and traffic management for in-vehicle service access.

[0027] (3) Expand the monitoring indicators of the service grid to cover vehicle-specific service quality and safety requirements, such as real-time performance, reliability, and functional safety.

[0028] Step 2: Deploy the automotive software anomaly detection module in the automotive service mesh control plane to subscribe to and analyze service runtime telemetry data in real time for real-time anomaly detection.

[0029] It should be noted that the automotive software anomaly detection module is deployed in the service grid control plane to subscribe to and analyze the telemetry data of the service runtime in real time, including integrating the anomaly detection service into the grid console, performing corresponding tests and optimizing the detection algorithm, etc. Combined with mechanisms such as on-board diagnostics (OBD) and fault injection, the anomaly detection capability is continuously evaluated and improved.

[0030] The specific implementation process includes:

[0031] (1) Define the anomaly detection service interface: First, you need to define a standardized anomaly detection service interface to clarify the service’s input and output data formats and interaction methods.

[0032] (2) Develop anomaly detection services: Based on the defined interfaces, develop microservices that implement anomaly detection functions. This includes implementing encapsulated anomaly detection algorithms, encapsulating deep learning anomaly detection algorithms based on multi-knowledge fusion as services, implementing data subscription and caching, implementing the gRPC or REST API interface of anomaly detection services, receiving telemetry data from the control plane, and returning anomaly detection results.

[0033] (3) Integrate the developed anomaly detection service into the service mesh control plane.

[0034] (4) For the service grid of automotive software, customize the telemetry data required for anomaly detection services, including: service call topology data, service input and output data, and resource usage data.

[0035] (5) Provide a management interface for anomaly detection services for automotive software development and operation and maintenance teams, integrated into the service mesh console.

[0036] (6) Perform functional and performance testing of anomaly detection services in a service mesh environment.

[0037] (7) Based on the operational data accumulated by the service grid, the anomaly detection model is continuously trained and optimized.

[0038] Specifically, the anomaly detection model F takes as input the telemetry data x of the automotive software service grid and outputs the anomaly determination result y.

[0039] y=F(x)

[0040] Among them, x contains monitoring indicator data of multiple services s at multiple time steps t, such as:

[0041] x={x_s1_t1,x_s1_t2,...,x_s1_tn,x_s2_t1,x_s2_t2,...,x_s2_tn,...,x_sm_t1,x_sm_t2,...,x_sm_tn}

[0042] Among them, model F includes: graph convolutional network (GCN), recurrent neural network (RNN) and attention mechanism.

[0043] Among them, the graph convolutional network is used to extract the structural features of the service call topology graph:

[0044] h_graph = GCN(A,X);

[0045] Among them, A is the adjacency matrix of the service call directed graph, X is the node feature matrix, GCN can be stacked in multiple layers, and the calculation formula for each layer is:

[0046] H^(l+1)=σ(D^-1 / 2·A·D^-1 / 2·H^(l)·W^(l))

[0047] Among them, D is the degree matrix of A, W is the weight matrix, and σ is the activation function (such as ReLU).

[0048] Recurrent neural networks are used to extract time series features of each service:

[0049] h_s1_rnn=RNN(x_s1_t1,x_s1_t2,...,x_s1_tn)

[0050] h_s2_rnn=RNN(x_s2_t1,x_s2_t2,...,x_s2_tn) ...

[0052] h_sm_rnn=RNN(x_sm_t1,x_sm_t2,...,x_sm_tn)

[0053] The definition of RNN is:

[0054] h_t=tanh(W_xh·x_t+W_hh·h_(t-1)+b_h)

[0055] The outputs of GCN and RNN are fused through the attention mechanism to generate the final anomaly score:

[0056] h_final=Attention(h_graph,h_s1_rnn,h_s2_rnn,...,h_sm_rnn)

[0057] y=Sigmoid(MLP(h_final))On this basis, an end-to-end deep learning algorithm is introduced to realize the upgrade of anomaly detection mode from local single data to global multimodal data.

[0058] The construction process of the anomaly detection model is as follows:

[0059] (1) Perform preprocessing operations such as cleaning, conversion, and normalization on the telemetry data collected by the automotive service grid. Specifically, remove noise data such as missing values ​​and outliers to ensure data quality, convert unstructured or semi-structured data (such as call links, logs, etc.) into structured feature vectors, and normalize the numerical ranges of different indicators to eliminate the impact of dimensions.

[0060] (2) Based on the preprocessed data, a directed graph model of the calling relationship between services is constructed. Each service is first abstracted as a node in the graph, and the node attributes include the service name, the management domain to which it belongs, etc. Then, directed edges are generated based on the calling relationship between services. As the calling relationship in the service grid changes dynamically, the nodes and edges of the graph are updated in real time.

[0061] (3) Use the Graph Convolutional Network (GCN) to extract features from the service call topology graph (a dynamic description of the vehicle service grid, which uses the process of data preprocessing - building a service call directed graph - using the graph convolutional network to extract features - generating anomaly detection models - model training and verification). First, learn the low-dimensional embedding representation of each service node through random walks or node attribute aggregation. Based on the node embedding, aggregate the feature information of neighboring nodes through graph convolution operations, update the node representation, and aggregate the node representation through graph pooling operations to generate the embedding vector of the entire service call topology graph.

[0062] (4) Recurrent neural networks are used to extract time series features of each service. Time series data includes: request data, response time data, error rate data, resource usage data, and call relationship data. Recurrent neural networks (such as LSTM and GRU) are used to extract features from these time series data. These neural network structures are suitable for processing data with time series relationships and can capture dependencies over long time spans. They can also analyze service performance, identify potential anomalies, and support subsequent anomaly detection models. Recurrent neural networks (such as LSTM and GRU) are used for feature extraction.

[0063] (5) Based on the features extracted by the graph convolutional network and the recurrent neural network, a deep learning anomaly detection model is designed. Specifically, the service call topology graph feature vector extracted by the graph convolutional network is v_graph, and the service time series feature vectors extracted by the recurrent neural network are v_s1_rnn, v_s2_rnn, ..., v_sm_rnn.

[0064] Feature fusion: concatenate graph features and time series features to form a long vector:

[0065] v_concat=Concat(v_graph,v_s1_rnn,v_s2_rnn,...,v_sm_rnn)

[0066] Among them, Concat is a tensor concatenation operation.

[0067] Feature transformation: Use the fully connected layer (Dense) to perform nonlinear transformation on the concatenated feature vectors to extract high-order interactive features:

[0068] h_fuse=ReLU(Dense(v_concat))

[0069] Among them, Dense is a fully connected layer and ReLU is an activation function, which can increase the nonlinear expression ability of the model. In order to further improve the quality of feature representation, multiple fully connected layers can be stacked.

[0070] h_fuse=ReLU(Dense(ReLU(Dense(v_concat))))

[0071] Anomaly score calculation: Finally, a Sigmoid activation function is connected to map the feature representation to between 0 and 1 to obtain the anomaly probability score:

[0072] y = Sigmoid(Dense(h_fuse))

[0073] Among them, the Sigmoid function compresses real numbers into the range of 0-1, and the closer to 1, the greater the possibility of abnormality.

[0074] In summary, by combining local feature extraction and end-to-end anomaly detection model, the complete forward calculation process can be expressed as:

[0075] v_graph = GCN(A,X)

[0076] v_s1_rnn=RNN(x_s1_t1,x_s1_t2,...,x_s1_tn)

[0077] v_s2_rnn=RNN(x_s2_t1,x_s2_t2,...,x_s2_tn) ...

[0079] v_sm_rnn=RNN(x_sm_t1,x_sm_t2,...,x_sm_tn)

[0080] v_concat=Concat(v_graph,v_s1_rnn,v_s2_rnn,...,v_sm_rnn)

[0081] h_fuse=ReLU(Dense(ReLU(Dense(v_concat))))

[0082] y=Sigmoid(Dense(h_fuse)).

[0083] It should be noted that the attention mechanism is designed at the high level of the deep anomaly detection model to adaptively learn the importance weights of different telemetry indicators for anomaly judgment. At the same time, the gating mechanism is introduced to enhance the model's memory and noise resistance. The specific implementation process includes:

[0084] (1) An attention mechanism is introduced at the high level of the anomaly detection model to adaptively learn the importance weights of different telemetry indicators.

[0085] (2) A gating mechanism is introduced at the high level of the anomaly detection model to enhance the model’s memory and noise resistance. The forgetting gate coefficient f is calculated through the sigmoid activation function to control the degree of retention of the memory state of the previous moment.

[0086] (3) Integrate the attention mechanism and gating mechanism into the end-to-end anomaly detection model, and use service mesh telemetry data for model training and optimization.

[0087] (4) Analyze the decision-making process and key influencing factors of the anomaly detection model.

[0088] In addition, the method of this application also includes using the service grid monitoring historical data accumulated by the automobile manufacturer to pre-train the anomaly detection model, and then using the real-time data of the fleet for continuous online learning to continuously improve the model performance and generalization ability.

[0089] The specific implementation process includes:

[0090] (1) Historical data collection and processing: Collect historical telemetry data within a certain time range (e.g., the past year) from the automobile manufacturer’s service grid monitoring system, including service call topology, performance indicators, resource utilization, etc., perform feature extraction and selection on the historical data after preprocessing operations such as cleaning, conversion, and normalization, and construct the input feature vector of the anomaly detection model. Features include the number of service calls, average latency, error rate, etc.

[0091] (2) Anomaly detection model pre-training: Based on the characteristics of the automotive software service grid, the structure and parameters of the anomaly detection model are initialized, and the anomaly detection model is trained using the training set data to optimize the model parameters. The performance of the pre-trained model is then evaluated on the validation set or test set. The performance of the model on the anomaly detection task is comprehensively measured by calculating indicators such as precision, recall, and F1 value. The advantages and disadvantages of the model and the direction of improvement are analyzed.

[0092] (3) Online learning of real-time fleet data: Through on-board terminal equipment and fleet management platforms: real-time collection of telemetry data of the vehicle service grid, including vehicle status data (reflecting the working status of each subsystem and component of the vehicle, such as the real-time operating parameters of the power system, braking system, steering system, battery management system, etc.), service call data (reflecting the call relationship and interaction between vehicle software services), environmental parameter data (external environment conditions of the vehicle, such as temperature, light intensity, road conditions, etc.), and vehicle driving data (driving status of the vehicle, such as speed, acceleration, steering angle, positioning information, etc.).

[0093] (4) Real-time detection: The real-time feature vector is input into the pre-trained anomaly detection model, the anomaly score is calculated, and whether an anomaly occurs is determined in real time based on the set threshold, generating an anomaly alert.

[0094] (5) Incremental learning and updating of models: New samples in the real-time data of the fleet (such as new service call patterns, abnormal events, etc.) are cached in memory or database as the data source for incremental learning. After incremental learning, the performance of the updated anomaly detection model is evaluated using a new validation set. If the performance is significantly improved, the updated model parameters are persisted to replace the original pre-trained model.

[0095] The pre-trained anomaly detection model is F_pre, and its parameter is θ_pre. The input of the model is the telemetry data x of the automotive software service grid, and the output is the anomaly judgment result y.

[0096] y = F_pre(x;θ_pre)

[0097] Among them, x can be expressed as a multi-dimensional feature vector of different time steps t and different services s, such as:

[0098] x={x_s1_t1,x_s1_t2,...,x_s1_tn,x_s2_t1,x_s2_t2,...,x_s2_tn,...,x_sm_t1,x_sm_t2,...,x_sm_tn}

[0099] According to the above content, the pre-training anomaly detection model F_pre is mainly composed of a graph convolutional network (GCN), a recurrent neural network (RNN) and a fully connected layer (Dense). The goal of the pre-training stage is to learn a set of initial parameters θ_pre with good generalization performance, laying the foundation for subsequent incremental learning.

[0100] In the incremental learning phase, new telemetry data x_new and corresponding anomaly labels y_new are obtained. At this time, the fine-tuning method can be used to continue training based on the pre-trained model and update the model parameters to adapt to the new data distribution. The updated model F_inc can be expressed as:

[0101] y_new = F_inc(x_new;θ_inc)

[0102] Among them, θ_inc is the new parameter obtained by incremental update based on θ_pre. The goal of incremental training is to minimize the loss function L on the new data, the cross entropy loss:

[0103] L(θ_inc)=-Σ[y_new·log(F_inc(x_new;θ_inc))+(1-y_new)·log(1-F_inc(x_new;

[0104] θ_inc))]

[0105] Use the gradient descent method to optimize the loss function and update the model parameters:

[0106]

[0107] The optional implementation modes of the embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation modes. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical scheme of the embodiments of the present invention, and these simple modifications all belong to the protection scope of the embodiments of the present invention.

Claims

1. A deep learning detection method for automotive software anomalies based on a service grid architecture, characterized in that: The steps include: Step 1: Build a lightweight automotive service grid covering the automotive electronic and electrical architecture to form a centrally managed data and control plane; Step 2: Deploy the automotive software anomaly detection module in the automotive service mesh control plane to subscribe to and analyze service runtime telemetry data in real time for real-time anomaly detection.

2. The method for detecting abnormalities in automotive software based on a service grid architecture according to claim 1, characterized in that: The service runtime telemetry data includes: service call topology data, service input and output data, and resource usage data.

3. The method for detecting abnormalities in automotive software based on a service grid architecture according to claim 2, characterized in that: The automobile software anomaly detection module includes an end-to-end deep learning anomaly detection model, and uses the anomaly detection model to perform real-time anomaly detection.

Citation Information

Patent Citations

  • Fault root cause positioning method and system based on service code level and storage medium

    CN113900844A

  • Intelligent optimization management and control method for equipment-intensive discrete type manufacturing factory

    CN116976549A

  • Database grid device realized based on sidecar technology and data processing method

    CN117370427A

  • Micro-service anomaly detection method and device based on graph attention network

    CN117407256A

  • Internet of vehicles real-time intrusion detection method, device and system

    CN117560183A