On-line diagnosis method, device and equipment for automobile audio bus and medium

Abstract

The invention provides an on-line diagnosis method, device and equipment for an automobile audio bus and a medium, the method is applied to a host, and the method comprises the following steps: generating a digital chaotic sequence different from a transmission signal in the automobile audio bus; injecting a digital chaos sequence into the bus to send the digital chaos sequence to a target child node in an idle time slot when the bus periodically transmits the transmission signal; receiving a response signal returned by the analog switch through the automobile audio bus; taking the response signal as a target chaotic sequence, and performing comparative analysis on the target chaotic sequence and the digital chaotic sequence; and determining whether the impedance mismatching problem exists on the communication path from the host to the child node and the analog switch based on the comparison analysis result. By means of the method, the technical problem that in the prior art, on the premise that normal communication of the bus is not interrupted, real-time, intelligent and automatic diagnosis and positioning cannot be conducted on the bus impedance state can be solved.

Description

Technical Field

[0001] This application relates to the field of automotive electronics and bus fault prediction technology, and more specifically, to an online diagnostic method, apparatus, device, and medium for automotive audio buses. Background Technology

[0002] In automotive electronic architecture, audio buses (such as A2B buses) are responsible for transmitting high-fidelity, low-latency audio data streams between the vehicle's head unit and multiple distributed audio modules (such as speakers and microphones). To ensure that signals are transmitted over long distances and through multiple nodes in the bus network without distortion or reflection, matching the characteristic impedance of the bus is crucial. Impedance matching is the physical basis for ensuring signal integrity and maintaining communication reliability.

[0003] Throughout a vehicle's lifecycle, factors such as vibration, temperature and humidity changes, connector aging, or accidental damage can cause localized impedance variations in the bus lines, resulting in impedance mismatch points. This mismatch can lead to signal reflection, consequently causing audio quality degradation, increased data error rates, and even failure of critical audio functions. Therefore, real-time monitoring and fault warning of the automotive audio bus's impedance matching status are crucial for improving system reliability and enabling predictive maintenance.

[0004] Currently, the industry primarily relies on offline testing equipment, such as time-domain reflectometers, for bus impedance testing. These methods require disconnecting the bus from the normal operating system, injecting a test signal, and analyzing the reflected waveform. This offline testing approach cannot be performed while the vehicle is running, cannot capture intermittent faults in dynamic changes, and is complex to operate, making it unsuitable for routine vehicle maintenance and health management.

[0005] Therefore, a solution is needed. Summary of the Invention

[0006] In view of this, embodiments of this application provide an online diagnostic method, apparatus, device, and medium for automotive audio buses to solve the technical problem that existing technologies cannot perform real-time, intelligent, and automated diagnosis and location of bus impedance status without interrupting normal bus communication.

[0007] In a first aspect, embodiments of this application provide an online diagnostic method for an automotive audio bus, applied to a host computer, the method comprising: Generate a digital chaotic sequence that is different from the current transmitted signal in the car audio bus; During the idle time slot when the transmission signal is periodically transmitted on the car audio bus, the digital chaotic sequence is injected into the car audio bus to be sent to the target sub-node. The system receives a response signal transmitted back from an analog switch connected to the target sub-node via the automotive audio bus; wherein the analog switch is configured to transmit a signal forwarded from the sub-node back to the host, and the target sub-node forwards the digital chaotic sequence to the analog switch after determining that it cannot recognize the digital chaotic sequence. The received response signal is used as the target chaotic sequence and compared with the digital chaotic sequence for analysis. Based on the comparative analysis results, it is determined whether there is an impedance mismatch problem in the communication path from the host to the child node and the analog switch.

[0008] In one feasible implementation, the step of comparing and analyzing the received response signal as a target chaotic sequence with the digital chaotic sequence includes: The target chaotic sequence is preprocessed in the time domain and transformed in the frequency domain to obtain its corresponding time domain data and frequency domain data; The time-domain data and the frequency-domain data are fused to generate a multi-dimensional feature vector characterizing the properties of the target chaotic sequence. The comparison analysis results are obtained based on the multidimensional feature vector and the digital chaotic sequence.

[0009] In a feasible implementation, obtaining the comparison analysis results based on the multidimensional feature vector and the digital chaotic sequence includes: The multidimensional feature vector and the digital chaotic sequence are input into a pre-trained neural network diagnostic model; Obtain the comparison analysis results output by the model. The comparison analysis results include: the result of whether the impedance is matched, the type of impedance mismatch, and / or the location range of the impedance mismatch.

[0010] In one feasible implementation, the neural network diagnostic model is a hybrid model combining a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to extract spatial features from the multidimensional feature vector, and the long short-term memory network is used to analyze the time-series dependence of signal changes.

[0011] In one feasible implementation, the method further includes: After determining that there is an impedance mismatch problem on the communication path based on the comparison analysis results, the comparison analysis results are sent to the active impedance compensation circuit connected to the car audio bus, so that the active impedance compensation circuit can compensate the impedance of the car audio bus in real time by adjusting its own variable capacitor and inductor array based on a preset compensation strategy.

[0012] In one feasible implementation, the method further includes: The comparison analysis results, injection parameters, and corresponding bus status information are stored in the historical database, and the neural network diagnostic model is retrained based on the updated historical data to optimize the neural network diagnostic model.

[0013] In one feasible implementation, the digital chaotic sequence is generated based on a Logistic map or a Lorentz system model, and before being injected into the automotive audio bus, the digital chaotic sequence is modulated to continuously cover a frequency band from 1 MHz to 28 MHz.

[0014] Secondly, embodiments of this application also provide an online diagnostic device for an automotive audio bus, mounted on a host computer, the device comprising: The sequence generation module is used to generate a digital chaotic sequence that is different from the current transmitted signal in the car audio bus; A signal injection module is used to inject the digital chaotic sequence into the car audio bus during the idle time slots when the transmission signal is periodically transmitted on the car audio bus, so as to send it to the target sub-node. A signal receiving module is used to receive a response signal transmitted back by an analog switch connected to the target sub-node through the car audio bus; wherein the analog switch is configured to transmit the signal forwarded from the sub-node back to the host, and the target sub-node forwards the digital chaotic sequence to the analog switch after determining that it cannot recognize the digital chaotic sequence. The signal comparison module is used to compare and analyze the received response signal as the target chaotic sequence with the digital chaotic sequence. The problem identification module is used to determine, based on the comparison analysis results, whether there is an impedance mismatch problem on the communication path from the host to the child node and the analog switch.

[0015] In one feasible implementation, the signal comparison module is used to compare and analyze the received response signal as a target chaotic sequence with the digital chaotic sequence, for the following purposes: The target chaotic sequence is preprocessed in the time domain and transformed in the frequency domain to obtain its corresponding time domain data and frequency domain data; The time-domain data and the frequency-domain data are fused to generate a multi-dimensional feature vector characterizing the properties of the target chaotic sequence. The comparison analysis results are obtained based on the multidimensional feature vector and the digital chaotic sequence.

[0016] In one feasible implementation, the signal comparison module is used to obtain the comparison analysis result based on the multidimensional feature vector and the digital chaotic sequence, for the following purposes: The multidimensional feature vector and the digital chaotic sequence are input into a pre-trained neural network diagnostic model; Obtain the comparison analysis results output by the model. The comparison analysis results include: the result of whether the impedance is matched, the type of impedance mismatch, and / or the location range of the impedance mismatch.

[0017] In one feasible implementation, the neural network diagnostic model is a hybrid model combining a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to extract spatial features from the multidimensional feature vector, and the long short-term memory network is used to analyze the time-series dependence of signal changes.

[0018] In one feasible implementation, the device further includes: The compensation module is used to send the comparison analysis results to the active impedance compensation circuit connected to the car audio bus after determining that there is an impedance mismatch problem on the communication path based on the comparison analysis results. This allows the active impedance compensation circuit to perform real-time compensation of the impedance of the car audio bus by adjusting its own variable capacitor and inductor array based on a preset compensation strategy.

[0019] In one feasible implementation, the device further includes: The storage module is used to store the comparison analysis results, injection parameters and corresponding bus status information to the historical database, and to retrain the neural network diagnostic model based on the updated historical data to optimize the neural network diagnostic model.

[0020] In one feasible implementation, the digital chaotic sequence is generated based on a Logistic map or a Lorentz system model, and before being injected into the automotive audio bus, the digital chaotic sequence is modulated to continuously cover a frequency band from 1 MHz to 28 MHz.

[0021] Thirdly, embodiments of this application also provide an electronic device, including: a processor, a storage medium, and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the online diagnostic method as described in any one of the first aspects.

[0022] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the online diagnostic method as described in any one of the first aspects.

[0023] This application provides an online diagnostic method, apparatus, device, and medium for an automotive audio bus. The method cleverly utilizes the inherent silence period of the system by injecting a digital chaotic sequence different from the currently transmitted signal into the idle time slot of the bus's periodic communication, thereby achieving the "invisible" implantation of the detection signal without interfering with the normal audio data transmission.

[0024] Furthermore, this embodiment creatively utilizes the target child node and its connected analog switch to form a closed bus loop. Specifically, the child node forwards unrecognizable detection signals to the analog switch, which then transmits the signal back to the host via the bus. This design allows the injected detection signal to undergo a complete transmission path on the bus from the host to the child node and then back to the host via the analog switch.

[0025] By comparing and analyzing the transmitted chaotic sequence with the received loop response signal, this valve stem can sense and assess changes in signal integrity throughout the entire closed communication path. Since any impedance mismatch will cause energy attenuation or waveform distortion during signal circulation, this change will be directly reflected in the return signal.

[0026] Compared to existing technologies that require interrupting bus communication and performing offline testing, this solution is the first to achieve online, non-intrusive diagnostics of the impedance matching status of the automotive audio bus. It allows for continuous and automatic monitoring of the bus's health status and fault warnings during normal vehicle operation without disconnecting the bus from the system. This fundamentally solves the technical challenges of offline testing methods, such as their inability to capture dynamic faults in real time, cumbersome operation, and disruption to continuous system operation.

[0027] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0028] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1A flowchart illustrating an online diagnostic method for an automotive audio bus provided in an embodiment of this application is shown.

[0030] Figure 2 This illustration shows a hardware structure diagram of an online diagnostic method for an automotive audio bus provided in an embodiment of this application.

[0031] Figure 3 This illustration shows a schematic diagram of the structure of an online diagnostic device for an automotive audio bus provided in an embodiment of this application.

[0032] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0034] In automotive electronic architecture, audio buses (such as A2B buses) are responsible for transmitting high-fidelity, low-latency audio data streams between the vehicle's head unit and multiple distributed audio modules (such as speakers and microphones). To ensure that signals are transmitted over long distances and through multiple nodes in the bus network without distortion or reflection, matching the characteristic impedance of the bus is crucial. Impedance matching is the physical basis for ensuring signal integrity and maintaining communication reliability.

[0035] Throughout a vehicle's lifecycle, factors such as vibration, temperature and humidity changes, connector aging, or accidental damage can cause localized impedance variations in the bus lines, resulting in impedance mismatch points. This mismatch can lead to signal reflection, consequently causing audio quality degradation, increased data error rates, and even failure of critical audio functions. Therefore, real-time monitoring and fault warning of the automotive audio bus's impedance matching status are crucial for improving system reliability and enabling predictive maintenance.

[0036] Currently, the industry primarily relies on offline testing equipment, such as time-domain reflectometers, for bus impedance testing. These methods require disconnecting the bus from the normal operating system, injecting a test signal, and analyzing the reflected waveform. This offline testing approach cannot be performed while the vehicle is running, cannot capture intermittent faults in dynamic changes, and is complex to operate, making it unsuitable for routine vehicle maintenance and health management.

[0037] Based on this, embodiments of this application provide an online diagnostic method, apparatus, device, and medium for automotive audio buses, which are described below through embodiments.

[0038] To facilitate understanding of this embodiment, a detailed description of an online diagnostic method for an automotive audio bus disclosed in this application embodiment will be provided first. This method is applied to a host computer. Figure 1 As shown, the method includes the following steps: Step 101: Generate a digital chaotic sequence that is different from the current transmitted signal in the car audio bus.

[0039] The method is executed by the head unit of the vehicle audio system. Here, the head unit typically refers to the central control unit of the audio system, which is responsible for managing the distribution of audio data streams, scheduling bus communications, and executing the diagnostic procedures described in this method. The head unit has a built-in or connected dedicated signal generation and processing module.

[0040] In this step, the host generates a digital chaotic sequence for detection. The key characteristic of this sequence is its significant difference in signal characteristics compared to the audio data stream currently being transmitted on the automotive audio bus (e.g., the A2B bus). This difference can be multifaceted; for example, in terms of spectral distribution, the energy of the detection sequence may more evenly cover the entire frequency band of the bus operation, while the audio data transmission signal may be concentrated near a specific carrier frequency. Alternatively, in terms of waveform characteristics, the sequence may exhibit noisy, non-periodic chaotic properties, distinctly different from the digital audio envelope with its specific frame structure.

[0041] The choice of using a digital chaotic sequence, distinct from normal transmission signals, as the detection stimulus is based on a thorough consideration of the physical layer characteristics of the A2B bus. The A2B bus is a differential communication system based on defined time slots, sensitive to signal edges and common-mode noise. If a fast-edge pulse signal is used as the detection signal, its steep edges are prone to complex reflections due to impedance mismatch, making analysis difficult. Furthermore, if the injection timing or method is inappropriate, the differential level may be misjudged as the start bit of a valid data frame, thus interfering with communication. In contrast, the chaotic sequence used in this embodiment is closer to broadband noise in waveform, exhibiting noise-like voltage fluctuations on the differential lines rather than logically significant "0" or "1" level transitions. Therefore, it can be effectively identified as "non-communication data" by the receiving node on the bus, thus being safely ignored or processed according to a preset path, fundamentally avoiding the risk of interfering with normal communication.

[0042] For example, the digital chaotic sequence is generated based on a Logistic map or a Lorentz system model, and before being injected into the automotive audio bus, the digital chaotic sequence is modulated so that its spectrum continuously covers a frequency band from 1 MHz to 28 MHz.

[0043] In other words, one feasible way to generate such sequences is to use known mathematical models of chaotic systems, such as the Logistic map or the Lorentz equation, and generate them through iterative calculations with initial parameters set. The aim is that when this sequence is injected into the bus, it can be clearly distinguished from normal communication data by the receiving nodes on the bus, thus avoiding misprocessing. Simultaneously, its wideband or specific characteristics help to stimulate the bus link's response at different frequencies, providing rich information for subsequent analysis. The generation operation can be performed in the host processor or a dedicated programmable logic device, and the generated sequence is temporarily stored in digital form.

[0044] Step 102: During the idle time slot when the transmission signal is periodically transmitted on the car audio bus, the digital chaotic sequence is injected into the car audio bus to be sent to the target sub-node.

[0045] Automotive audio buses, such as the A2B bus, typically use communication protocols that are scheduled based on defined, periodic time slots. Within each communication cycle, a portion of the time slots are pre-allocated for transmitting valid audio data packets, while there are brief, unused idle periods between data packets or at specific stages of the communication cycle. This step utilizes these protocol-inherent, predictable quiet intervals to perform the detection operation.

[0046] Specifically, the host precisely controls the timing of the detection signal injection based on its understanding of the bus communication protocol. Within a communication frame, it selects a time window outside of the time slots allocated specifically for audio data transmission. For example, within a 10-millisecond communication cycle, it might select a guard band or inter-frame gap lasting approximately 100 microseconds, which does not carry any service data. During this selected, brief idle time slot, the host applies the previously generated digital chaotic sequence as an electrical signal to the physical lines of the bus via its internal drive circuitry or an external signal coupling unit.

[0047] This design ensures that the injection of the detection signal is completely time-staggered from normal audio data transmission. Therefore, the detection signal does not consume bandwidth or interfere with or conflict with the transmitting audio data packets, achieving true "online" and "disruptive" detection. The injected signal propagates along the bus to all connected child nodes, and its intended destination, depending on the configuration, is a specific target child node.

[0048] Step 103: Receive the response signal transmitted back by the analog switch connected to the target sub-node through the car audio bus; wherein the analog switch is configured to transmit the signal forwarded from the sub-node back to the host, and the target sub-node forwards the digital chaotic sequence to the analog switch after determining that it cannot recognize the digital chaotic sequence.

[0049] After the detection signal injected in the previous step propagates along the bus to the target child node, the receiver of the target child node will receive the signal. Since this signal is designed to be inconsistent with the data frame format of the standard communication protocol, the protocol processing unit of the child node will determine it as "unrecognizable" or "non-target data." In some implementations, this determination can be based on hardware signal feature matching circuitry, for example, detecting that the signal lacks a valid frame header; in other designs, it can also be determined by firmware performing simple pattern matching.

[0050] The key is that the target sub-node is pre-configured or designed to not simply discard such unrecognizable signals, but instead initiate a special forwarding process. Internally or via an external connection, the target sub-node has a path that guides the electrical representation (or buffered and restructured digital form) of the signal to a functional unit called an "analog switch." This analog switch can be a physical array of electronic switches or a logic module with signal feedback capabilities. Its core function is, when triggered, to re-drive the input signal from the target sub-node back onto the bus, either along its original path or via a specified path.

[0051] Therefore, after the target child node completes its forwarding action, the analog switch is activated. It then applies the received signal back to the car audio bus through its output drive circuit. At this point, the signal begins its return journey, propagating along the bus towards the host. Finally, at the host, the circuit responsible for signal acquisition (which can be a coupler shared with the injection circuit or a separate receiving unit) captures this signal that has completed a full "round trip" and converts it into a digital form suitable for subsequent processing, namely the "response signal" described in this step.

[0052] This mechanism cleverly constructs a "signal loop" with the host as the starting and ending point and the target sub-nodes and analog switches as intermediate nodes. The injected detection signal will propagate twice (outbound and return) along the bus path covered by this loop. Impedance anomalies at any point on the path will have a cumulative effect on the signal and will ultimately be reflected in the response signal received by the host.

[0053] Step 104: The received response signal is used as the target chaotic sequence and compared with the digital chaotic sequence.

[0054] After receiving the response signal from the analog switch, the host obtains two sets of related signal data: one is the original, pure reference signal, i.e., the initially injected digital chaotic sequence; the other is the response signal carrying path state information, returned after completing the full bus loop transmission, which can be referred to here as the target chaotic sequence. The core of this step lies in deeply comparing these two sequences to extract the fault information contained in the differences.

[0055] Comparative analysis is not simply a direct subtraction or comparison. Its purpose is to quantify or characterize the changes in the response signal relative to the original signal. These changes may include overall amplitude attenuation, enhancement or attenuation of specific frequency components, time broadening or distortion of the waveform, and reduction in sequence correlation. Each pattern of change may be associated with a specific type and location of impedance anomaly in the loop.

[0056] To achieve effective analysis, various signal processing methods can be employed. For example, the cross-correlation function of two sequences can be calculated, and signal integrity can be assessed by observing the amplitude attenuation and width variation of the correlation peaks. Alternatively, time-frequency analysis can be performed on the two sequences separately to extract their respective time-domain statistical features (such as mean, variance, and kurtosis) and frequency-domain features (such as spectral energy distribution and main peak frequency). Then, by comparing the differences in these features, an index characterizing signal distortion can be constructed. These processing methods aim to transform the impact of loop transmission on the signal into one or more sets of quantitative features that can be used for subsequent intelligent judgment.

[0057] Through this comparative analysis process, the host transforms the complex effects of the bus, a physical transmission medium, into data characteristics that can be calculated and processed logically, laying the foundation for the final determination of impedance matching status.

[0058] Step 105: Based on the comparison analysis results, determine whether there is an impedance mismatch problem on the communication path from the host to the child node and the analog switch.

[0059] After completing signal comparison and extracting the corresponding features or indicators, the core task of this step is to determine the health status of the bus loop that the signal traverses based on these analysis results. Specifically, it involves determining whether there are impedance mismatch points that could cause signal reflection or distortion along the complete closed path from the host, through the bus to the target sub-node, and then back to the host via the analog switch.

[0060] The logic for making judgments can be based on preset thresholds or more complex decision-making models. For example, if comparative analysis shows that the overall signal energy attenuation exceeds a certain threshold, or the amplitude of the main cross-correlation peak drops below a certain level, it can be preliminarily determined that there is an impedance anomaly in the loop that causes significant energy loss, such as poor connection or a sharp increase in line loss. If the analysis finds that the signal waveform exhibits abnormal broadening or splitting, it may indicate the presence of distributed parameter changes that cause signal dispersion.

[0061] Through the above steps, this embodiment of the application completes a closed loop from "excitation injection" to "state determination", realizing online and uninterrupted health status diagnosis of specific paths of the automotive audio bus.

[0062] This application provides an online diagnostic method, apparatus, device, and medium for an automotive audio bus. The method cleverly utilizes the inherent silence period of the system by injecting a digital chaotic sequence different from the currently transmitted signal into the idle time slot of the bus's periodic communication, thereby achieving the "invisible" implantation of the detection signal without interfering with the normal audio data transmission.

[0063] Furthermore, this embodiment creatively utilizes the target child node and its connected analog switch to form a closed bus loop. Specifically, the child node forwards unrecognizable detection signals to the analog switch, which then transmits the signal back to the host via the bus. This design allows the injected detection signal to undergo a complete transmission path on the bus from the host to the child node and then back to the host via the analog switch.

[0064] By comparing and analyzing the transmitted chaotic sequence with the received loop response signal, this valve stem can sense and assess changes in signal integrity throughout the entire closed communication path. Since any impedance mismatch will cause energy attenuation or waveform distortion during signal circulation, this change will be directly reflected in the return signal.

[0065] Compared to existing technologies that require interrupting bus communication and performing offline testing, this solution is the first to achieve online, non-intrusive diagnostics of the impedance matching status of the automotive audio bus. It allows for continuous and automatic monitoring of the bus's health status and fault warnings during normal vehicle operation without disconnecting the bus from the system. This fundamentally solves the technical challenges of offline testing methods, such as their inability to capture dynamic faults in real time, cumbersome operation, and disruption to continuous system operation.

[0066] In one feasible implementation, the step of comparing and analyzing the received response signal as a target chaotic sequence with the digital chaotic sequence includes: The target chaotic sequence is preprocessed in the time domain and transformed in the frequency domain to obtain its corresponding time domain data and frequency domain data; the time domain data and the frequency domain data are fused to generate a multidimensional feature vector characterizing the characteristics of the target chaotic sequence; based on the multidimensional feature vector and the digital chaotic sequence, the comparison analysis result is obtained.

[0067] In this implementation, the analysis process first focuses on processing the returned response signal, i.e., the target chaotic sequence. In the time domain, a series of preprocessing operations are performed. For example, digital filters can be used to suppress high-frequency noise or irrelevant low-frequency interference in the signal, making the signal components of interest more prominent. Subsequently, the filtered data may be normalized to eliminate the amplitude effects caused by small fluctuations in injected power or changes in acquisition gain, normalizing the signal to a uniform scaling range for stable feature extraction later.

[0068] After time-domain preprocessing, the target chaotic sequence is transformed to the frequency domain for observation. A typical approach is to perform a Fast Fourier Transform (FFT), such as a 4096-point FFT, to obtain the energy distribution spectrum of the signal in the frequency dimension, i.e., frequency domain data. At this point, we have a representation of the same signal from two different perspectives: the time domain (waveform, statistics) and the frequency domain (spectrum, harmonics).

[0069] Next, instead of using time-domain or frequency-domain data in isolation, the information from both domains is fused. Specifically, this fusion can be achieved through algorithms such as principal component analysis (PCA) to extract the most important, non-redundant information components from the time-domain waveform features and frequency-domain spectral features, combining them into a structured dataset containing multi-dimensional characteristics of the signal—a multi-dimensional feature vector. This vector comprehensively characterizes the overall "appearance" of the received signal after it has undergone loop transmission.

[0070] Thus, the analysis yields a condensed feature profile describing the "result" (the target chaotic sequence). Information about the "cause" (the original injected sequence), i.e., the digital chaotic sequence itself or its known properties (such as theoretical spectrum and initial parameters), serves as another crucial input. Ultimately, the comparative analysis results are derived by comprehensively considering this multidimensional feature vector representing the result state, along with the prior knowledge inherent in the original digital chaotic sequence serving as a reference. For example, this feature vector, along with the known features of the original sequence, can be input into an intelligent analysis unit capable of recognizing complex patterns for joint judgment.

[0071] In one exemplary implementation, obtaining the comparison analysis result based on the multidimensional feature vector and the digital chaotic sequence includes: The multidimensional feature vector and the digital chaotic sequence are input into a pre-trained neural network diagnostic model; the comparison analysis results output by the model are obtained, including: the result of impedance matching, the type of impedance mismatch, and / or the location range of impedance mismatch.

[0072] In this implementation, a pre-trained neural network diagnostic model is used to make the final intelligent judgment. Specifically, the multi-dimensional feature vector generated in the previous steps, which comprehensively represents the characteristics of the current loop response signal, along with the original digital chaotic sequence (or its appropriately processed feature representation) serving as the excitation source reference information, is provided as input to this neural network model.

[0073] This neural network model performs in-depth processing of input information through a series of complex nonlinear calculations and feature transformations. It can automatically learn and understand: what characteristic pattern of response will be generated by a specific original excitation signal when the bus loop is in a healthy state; and how the characteristics of the response signal will systematically deviate from the normal pattern when there are different types (such as short circuit, open circuit, poor contact) or impedance anomalies in different sections of the loop.

[0074] After forward propagation calculations within the model, the final, definitive result of the comparison is obtained from its output layer. This result can be a comprehensive judgment, such as directly stating "impedance matching" or "impedance mismatch." Furthermore, the model can also output more detailed diagnostic information, such as specifying the exact type of impedance mismatch detected, inferring from signal distortion characteristics whether the fault point is located on the path from the host to the child node, or on the path from the analog switch back to the host, or providing a more specific location range estimate.

[0075] By introducing a pre-trained neural network model, the complex tasks of signal distortion pattern recognition and fault mapping are delegated to algorithms with powerful learning capabilities. This enables the diagnostic process to handle highly nonlinear and subtle fault conditions, reduces reliance on manually set fixed thresholds, and improves the automation, adaptability, and overall accuracy of the diagnostic process for different fault modes.

[0076] In one feasible implementation, the neural network diagnostic model is a hybrid model combining a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to extract spatial features from the multidimensional feature vector, and the long short-term memory network is used to analyze the time-series dependence of signal changes.

[0077] That is, in the embodiments of this application, the neural network diagnostic model is not a single type of network, but is an organic combination of two sub-networks with complementary functions to collaboratively process the input feature data.

[0078] The first part of the model employs a convolutional neural network (CNN). It acts as a sophisticated feature filter. When a multidimensional feature vector, incorporating information from both the time and frequency domains, is input, the CNN operates through its internal convolutional and pooling layers. These layers automatically and progressively extract discriminative local patterns and structural information from the input data. For example, it can identify peak combinations representing specific resonant frequencies or feature segments characterizing typical waveform distortions within the feature vector. This extracted information can be understood as the "spatial features" or "local patterns" exhibited by the signal in a multidimensional feature space.

[0079] The second part of the model employs a Long Short-Term Memory (LSTM) network. This part of the network is designed to handle dependencies related to sequences or time. It takes high-level features (or sequences thereof) extracted by a convolutional neural network as input. LSM networks, through their internal gating mechanisms, selectively remember important historical information and understand how these features evolve with processing steps (corresponding to logical steps or implicit time dimensions in signal analysis). This allows the model to not only capture a snapshot of fault features at a given moment but also understand the dynamic process and correlations of feature changes, such as analyzing the broadening effect or multipath interference patterns of reflected waves caused by a fault on the time axis.

[0080] In this hybrid model architecture, the convolutional neural network is responsible for extracting key clues from static, multi-dimensional feature "pictures," while the long short-term memory network is responsible for interpreting the dynamic connections and evolutionary logic between these clues. Working together, the two enable the model to more comprehensively and deeply understand the complex signal response patterns caused by the bus loop state, thus making more robust and accurate diagnoses.

[0081] By employing a hybrid model combining Convolutional Neural Networks (CNNs) and Long Short-Term Memory Networks (LSTMs), this approach leverages the strengths of CNNs in spatial feature extraction and the expertise of LSTMs in time series modeling. This architecture is particularly well-suited for handling the complex diagnostic problems presented in this application, which combine spatial patterns (such as spectral feature distribution) and temporal dynamics (such as reflected wave time series). It can more fully extract information from feature vectors and is expected to improve the diagnostic and differentiation capabilities for complex and compound faults.

[0082] In one feasible implementation, the method further includes: After determining that there is an impedance mismatch problem on the communication path based on the comparison analysis results, the comparison analysis results are sent to the active impedance compensation circuit connected to the car audio bus, so that the active impedance compensation circuit can compensate the impedance of the car audio bus in real time by adjusting its own variable capacitor and inductor array based on a preset compensation strategy.

[0083] When, based on the aforementioned comparative analysis, it is determined that an impedance mismatch problem does indeed exist on the communication path from the host to the child nodes and analog switches, the embodiments of this application do not stop at diagnosis but further perform active correction actions. That is, the diagnostic results, including information such as the fault type, possible location, or severity, are sent to a dedicated hardware unit connected to the automotive audio bus, known as an "active impedance compensation circuit."

[0084] Upon receiving a diagnostic command, this active impedance compensation circuit does not operate blindly, but rather works according to its internally preset or dynamically adjustable compensation strategy. The compensation strategy defines the specific corrective measures to be taken for different diagnostic results (e.g., a determination of high-frequency capacitive mismatch or low-frequency inductive mismatch).

[0085] The core physical mechanism of compensation is achieved by adjusting the component parameters of the variable capacitor and variable inductor arrays contained within the compensation circuit. For example, if a diagnostic indicates that the bus exhibits excessive capacitance at a specific frequency, the compensation strategy may instruct the circuit to increase the inductance of its inductor components to introduce an inductive component with opposite characteristics for neutralization; conversely, the same applies. These capacitors and inductors are coupled to the bus through an appropriate network topology, and changes in their parameters alter the overall equivalent impedance as seen from the bus.

[0086] Through this real-time, adjustable parameter adjustment, the active impedance compensation circuit can dynamically generate a compensation impedance opposite to the detected impedance anomaly and apply it to the bus. Its goal is to offset or mitigate impedance mismatches caused by line aging, connector problems, or external interference, thereby restoring or approaching the ideal matching state of the overall impedance characteristics of the bus system to maintain high-quality signal transmission.

[0087] This closed-loop integration of the diagnostic system and the actuator enables a leap from "fault detection" to "proactive repair." This not only promptly curbs signal quality degradation caused by impedance mismatch, preventing the escalation of fault impacts, but also significantly enhances the adaptive capability and long-term operational stability of the bus system, providing crucial technical support for predictive maintenance.

[0088] In one feasible implementation, the method further includes: The comparison analysis results, injection parameters, and corresponding bus status information are stored in the historical database, and the neural network diagnostic model is retrained based on the updated historical data to optimize the neural network diagnostic model.

[0089] In other words, a continuously growing knowledge base is built and maintained simultaneously during the diagnosis and compensation process. A series of key data generated from each complete diagnostic cycle—from injecting a chaotic sequence with specific parameters to finally obtaining the comparative analysis results—are systematically collected and stored. This data mainly includes: the specific parameters of the injected signal (e.g., the initial value of the chaotic mapping, control parameter μ), the characteristics of the acquired loop response signal, the detailed comparative analysis results output by the neural network model (e.g., diagnostic conclusions, confidence levels, fault characteristic values), and the known or presumed state of the bus during operation (e.g., ambient temperature, vehicle operating mode as context labels). All this information is linked together to form a complete historical record, stored in a dedicated historical database.

[0090] As vehicles continuously execute this diagnostic method during long-term operation, this historical database will accumulate massive amounts of real-world data covering various operating conditions, wiring aging stages, and different fault modes. Based on this increasingly rich database, the retraining or optimization process of the neural network diagnostic model can be initiated periodically or when specific conditions are met.

[0091] During this process, the latest historical data is merged with the existing data to form an updated training dataset. By re-executing the model training process using this data, the neural network can continuously learn the characteristic patterns under new operating conditions or for newly emerging failure modes. Through this iterative learning, the model's parameters and internal decision boundaries are optimized, and its diagnostic accuracy, adaptability to different degradation modes, and sensitivity to early, subtle faults are expected to be continuously improved.

[0092] By establishing a closed-loop feedback and learning mechanism for diagnostic data, it can possess the ability to self-evolve and optimize. It can not only accumulate valuable field data assets, but also make the core diagnostic model "smarter" as it is used over time, thereby continuously improving the reliability of long-term use and effectively dealing with the slow performance changes caused by the aging of the bus system. It is a key link in realizing intelligent, adaptive, and predictive maintenance.

[0093] Figure 2 This illustration shows a hardware structure diagram of an online diagnostic method for an automotive audio bus provided in an embodiment of this application. Figure 2 As shown, the core hardware architecture of this solution is an integrated, programmable signal processing and control system used to implement the aforementioned "online, disturbance-free, closed-loop diagnosis and compensation" method. The function of each module / device is analyzed below: The GPU is used to control the execution of the entire scheme: it is the "processing core" of the entire system.

[0094] DDR and FLASH: These are the system's "memory" units. DDR (memory) is used for high-speed caching of real-time acquired response signal data, intermediate processing results, and neural network model parameters, supporting high-speed computation by FPGA / GPU. FLASH (flash memory) is used for non-volatile storage, such as storing historical databases (diagnostic records, injected parameters, bus status), trained neural network diagnostic model firmware, preset compensation strategies, and system boot programs.

[0095] DAC conversion circuit: converts digital signals into analog signals.

[0096] Variable capacitor and inductor array: The physical actuator for "active compensation". This is an adjustable passive network controlled by an FPGA or microcontroller through a digital interface (such as I2C, SPI). Based on the fault type and location information output by the neural network diagnostic model, the FPGA calculates the optimal compensation impedance according to a preset compensation strategy, and adjusts the capacitance / inductance values ​​of each capacitor and inductor in the array through drive circuits (such as digital potentiometers, MEMS switches), connecting them in parallel or series to the bus, thereby correcting the detected impedance mismatch in real time.

[0097] CAN and LIN: "Information interfaces" for communication between the system and the vehicle. They are used to receive commands from the vehicle control unit (VCU) (such as initiating diagnostics), report diagnostic results (fault codes, locations, severity levels) to the instrument panel or cloud, and may receive contextual information from other sensors (such as vehicle status and ambient temperature) to enrich the bus status information and store it in the historical database.

[0098] The GPU controls the FPGA chaos algorithm generator to complete the generation, injection, reception, and comparison of digital chaotic sequences. The path is as follows: chaos algorithm generator — DAC conversion circuit — low-pass filter — operational amplifier circuit — high-speed isolation transformer — analog switch — broadband RF transformer — low-noise amplifier — ADC sampling circuit — time and frequency domain processing circuit.

[0099] In other words, the first step: generation and conditioning of the excitation signal (host side, signal transmission): FPGA Chaos Algorithm Generator: The source of digital chaotic sequences. Within the FPGA, it generates wideband, noise-like digital chaotic sequences through real-time iterative calculations using logic circuits (such as Logistic mapping). This serves as the digital prototype of the "probe" for the entire detection scheme.

[0100] DAC (Digital-to-Analog Converter) circuit: Converts the chaotic digital sequence generated by the FPGA into the corresponding analog voltage waveform. This serves as a bridge for signals to enter the analog physical world from the digital domain.

[0101] Low-pass filter (located after the DAC): Reconstruction filtering and band limiting. It filters out high-frequency image components and glitches above the Nyquist frequency generated during DAC conversion, smoothing the output analog signal and strictly limiting its spectrum within the target detection band (e.g., 1-28MHz) to prevent out-of-band noise from interfering with the bus.

[0102] Operational amplifier circuit (driver stage): Signal amplification and driving. It amplifies the filtered analog chaotic signal to a suitable amplitude and provides sufficient output current to drive the primary coil of the subsequent isolation transformer, ensuring that the signal can be effectively coupled out.

[0103] High-speed isolation transformer: Electrical isolation, single-ended to differential conversion, impedance matching / injection coupling. Electrical isolation: Completely isolates the host-side circuitry from the A2B bus, ensuring safety. Single-ended to differential conversion: Converts the single-ended signal from the preceding stage into the differential signal required by the A2B bus. Coupling injection: Injects the differential detection signal onto the A2B bus differential lines through transformer magnetic coupling in a high-impedance parallel connection. Its high-impedance characteristics ensure minimal impact on the original bus signals.

[0104] Step 2: Signal flow in the bus loop: A2B bus network: the transmission medium for detection signals and the object of diagnosis. Injected chaotic signals propagate along the bus.

[0105] Path A (outbound): The detection signal (digital chaotic sequence) starts from the host side and is transmitted to the target child node.

[0106] Target child node: Its receiver receives the signal, but the protocol stack determines it as a "non-data" frame.

[0107] Analog switch: Under the control of the target sub-node, this switch is activated to physically switch the received signal to the return path.

[0108] Path B (Return): After passing through the analog switch, the signal is reconnected / coupled back to the A2B bus and begins to return towards the host.

[0109] Step 3: Acquisition and digitization of response signals (host side, signal recovery): Broadband RF transformer (acquisition side): Couples and isolates the acquisition. The response signal returned from the bus (containing full impedance information of the loop path) is coupled and electrical isolation is provided again. It is typically physically independent of or highly directional from the isolation transformer on the injection side to avoid shoot-through signal interference.

[0110] Low-noise amplifier: Preamplifier. This stage amplifies the very weak response signal coupled from the transformer. This amplifier must have extremely low inherent noise to amplify the signal without introducing excessive additional noise, thus maintaining the original signal-to-noise ratio as much as possible.

[0111] ADC sampling circuit: Analog-to-digital conversion and synchronous acquisition. It synchronously converts the amplified analog response signal into a digital sequence at a high sampling rate (Super Nyquist). "Synchronization" here is crucial; it must maintain a strictly known timing relationship with the generation of the excitation signal to ensure subsequent coherent processing.

[0112] Step 4: Signal Analysis and Diagnosis (Digital Domain Processing): FPGA time-frequency domain processing circuit: Core digital signal processing. Processing the digital response sequence acquired by the ADC: Digital filtering: Further filters out out-of-band noise.

[0113] Time-frequency transformation: Performing FFT and other operations to transform the signal from the time domain to the frequency domain.

[0114] Feature extraction and fusion: Extract features from time-domain and frequency-domain data and fuse them into a multi-dimensional feature vector that characterizes signal distortion.

[0115] The processed feature vectors, along with the information from the original chaotic sequence, are fed into the neural network diagnostic model (also located within the FPGA) for intelligent diagnosis.

[0116] In this embodiment, an FPGA is used as the core for precise synchronization and control. During the quiet intervals of bus communication, a chaotic signal generation and injection link is driven to implant a broadband probe, which is fundamentally different from the business data, into the bus without interference through a high-speed isolation transformer. After traversing a predetermined loop composed of target child nodes and analog switches, the signal carries all impedance information along the path and is then sensitively retrieved by a low-noise amplifier and ADC through a high-fidelity acquisition link.

[0117] Ultimately, the signal undergoes full processing within the FPGA, from time-frequency domain feature extraction to neural network intelligent diagnosis, and can drive a variable capacitor-inductor array for real-time compensation. This hardware-software collaborative closed loop, from "digital generation" to "physical injection," then to "physical recycling," and finally back to "digital intelligent analysis," makes online, non-sensory, and adaptive monitoring and maintenance of bus impedance status possible, surpassing the fundamental limitations of traditional offline and disconnection detection methods.

[0118] Based on the same technical concept, this application also provides an online diagnostic device for an automotive audio bus, mounted on a host computer, such as... Figure 3 As shown, the device includes: Sequence generation module 301 is used to generate a digital chaotic sequence that is different from the current transmitted signal in the car audio bus.

[0119] The signal injection module 302 is used to inject the digital chaotic sequence into the car audio bus during the idle time slot when the transmission signal is periodically transmitted on the car audio bus, so as to send it to the target sub-node.

[0120] The signal receiving module 303 is used to receive the response signal transmitted back by the analog switch connected to the target sub-node through the car audio bus; wherein the analog switch is configured to transmit the signal forwarded from the sub-node back to the host, and after the target sub-node determines that it cannot recognize the digital chaotic sequence, it forwards the digital chaotic sequence to the analog switch.

[0121] The signal comparison module 304 is used to compare and analyze the received response signal as the target chaotic sequence with the digital chaotic sequence.

[0122] Problem determination module 305 is used to determine, based on the comparison analysis results, whether there is an impedance mismatch problem in the communication path from the host to the child node and the analog switch.

[0123] In one feasible implementation, the signal comparison module is used to compare and analyze the received response signal as a target chaotic sequence with the digital chaotic sequence, for the following purposes: The target chaotic sequence is preprocessed in the time domain and transformed in the frequency domain to obtain its corresponding time domain data and frequency domain data.

[0124] The time-domain data and the frequency-domain data are fused to generate a multi-dimensional feature vector characterizing the properties of the target chaotic sequence.

[0125] The comparison analysis results are obtained based on the multidimensional feature vector and the digital chaotic sequence.

[0126] In one feasible implementation, the signal comparison module is used to obtain the comparison analysis result based on the multidimensional feature vector and the digital chaotic sequence, for the following purposes: The multidimensional feature vector and the digital chaotic sequence are input into a pre-trained neural network diagnostic model.

[0127] Obtain the comparison analysis results output by the model. The comparison analysis results include: the result of whether the impedance is matched, the type of impedance mismatch, and / or the location range of the impedance mismatch.

[0128] In one feasible implementation, the neural network diagnostic model is a hybrid model combining a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to extract spatial features from the multidimensional feature vector, and the long short-term memory network is used to analyze the time-series dependence of signal changes.

[0129] In one feasible implementation, the device further includes: The compensation module is used to send the comparison analysis results to the active impedance compensation circuit connected to the car audio bus after determining that there is an impedance mismatch problem on the communication path based on the comparison analysis results. This allows the active impedance compensation circuit to perform real-time compensation of the impedance of the car audio bus by adjusting its own variable capacitor and inductor array based on a preset compensation strategy.

[0130] In one feasible implementation, the device further includes: The storage module is used to store the comparison analysis results, injection parameters and corresponding bus status information to the historical database, and to retrain the neural network diagnostic model based on the updated historical data to optimize the neural network diagnostic model.

[0131] In one feasible implementation, the digital chaotic sequence is generated based on a Logistic map or a Lorentz system model, and before being injected into the automotive audio bus, the digital chaotic sequence is modulated to continuously cover a frequency band from 1 MHz to 28 MHz.

[0132] This application provides an online diagnostic method, apparatus, device, and medium for an automotive audio bus. The method cleverly utilizes the inherent silence period of the system by injecting a digital chaotic sequence different from the currently transmitted signal into the idle time slot of the bus's periodic communication, thereby achieving the "invisible" implantation of the detection signal without interfering with the normal audio data transmission.

[0133] Furthermore, this embodiment creatively utilizes the target child node and its connected analog switch to form a closed bus loop. Specifically, the child node forwards unrecognizable detection signals to the analog switch, which then transmits the signal back to the host via the bus. This design allows the injected detection signal to undergo a complete transmission path on the bus from the host to the child node and then back to the host via the analog switch.

[0134] By comparing and analyzing the transmitted chaotic sequence with the received loop response signal, this valve stem can sense and assess changes in signal integrity throughout the entire closed communication path. Since any impedance mismatch will cause energy attenuation or waveform distortion during signal circulation, this change will be directly reflected in the return signal.

[0135] Compared to existing technologies that require interrupting bus communication and performing offline testing, this solution is the first to achieve online, non-intrusive diagnostics of the impedance matching status of the automotive audio bus. It allows for continuous and automatic monitoring of the bus's health status and fault warnings during normal vehicle operation without disconnecting the bus from the system. This fundamentally solves the technical challenges of offline testing methods, such as their inability to capture dynamic faults in real time, cumbersome operation, and disruption to continuous system operation.

[0136] Figure 4 A schematic diagram of an electronic device provided in this application embodiment includes: a processor 401, a storage medium 402, and a bus 403. The storage medium 402 stores machine-readable instructions executable by the processor 401. When the electronic device runs the online diagnostic method as described in the embodiment, the processor 401 communicates with the storage medium 402 via the bus 403, and the processor 401 executes the machine-readable instructions to perform the steps as described in the embodiment.

[0137] In this embodiment, the storage medium 402 may also execute other machine-readable instructions to perform other methods as described in the embodiment. For details on the specific execution steps and principles, please refer to the description of the embodiment, which will not be repeated here.

[0138] This application also provides a computer-readable storage medium storing a computer program that is executed by a processor to perform the steps as described in the embodiments.

[0139] In this embodiment, the computer program, when run by the processor, can also execute other machine-readable instructions to perform other methods as described in the embodiments. For details on the specific execution steps and principles, please refer to the description of the embodiments, which will not be repeated here.

[0140] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0141] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0142] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0143] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0144] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An online diagnostic method for an automotive audio bus, characterized in that, Applied to a host, the method includes: Generate a digital chaotic sequence that is different from the current transmitted signal in the car audio bus; During the idle time slot when the transmission signal is periodically transmitted on the car audio bus, the digital chaotic sequence is injected into the car audio bus to be sent to the target sub-node. The system receives a response signal transmitted back from an analog switch connected to the target sub-node via the automotive audio bus; wherein the analog switch is configured to transmit a signal forwarded from the sub-node back to the host, and the target sub-node forwards the digital chaotic sequence to the analog switch after determining that it cannot recognize the digital chaotic sequence. The received response signal is used as the target chaotic sequence and compared with the digital chaotic sequence for analysis. Based on the comparative analysis results, it is determined whether there is an impedance mismatch problem in the communication path from the host to the child node and the analog switch.

2. The method according to claim 1, characterized in that, The step of comparing and analyzing the received response signal as a target chaotic sequence with the digital chaotic sequence includes: The target chaotic sequence is preprocessed in the time domain and transformed in the frequency domain to obtain its corresponding time domain data and frequency domain data; The time-domain data and the frequency-domain data are fused to generate a multi-dimensional feature vector characterizing the properties of the target chaotic sequence. The comparison analysis results are obtained based on the multidimensional feature vector and the digital chaotic sequence.

3. The method according to claim 2, characterized in that, The comparison analysis results obtained based on the multidimensional feature vector and the digital chaotic sequence include: The multidimensional feature vector and the digital chaotic sequence are input into a pre-trained neural network diagnostic model; Obtain the comparison analysis results output by the model. The comparison analysis results include: the result of whether the impedance is matched, the type of impedance mismatch, and / or the location range of the impedance mismatch.

4. The method according to claim 3, characterized in that, The neural network diagnostic model is a hybrid model combining a convolutional neural network and a long short-term memory network. The convolutional neural network is used to extract spatial features from the multidimensional feature vector, and the long short-term memory network is used to analyze the time-series dependence of signal changes.

5. The method according to claim 1, characterized in that, The method further includes: After determining that there is an impedance mismatch problem on the communication path based on the comparison analysis results, the comparison analysis results are sent to the active impedance compensation circuit connected to the car audio bus, so that the active impedance compensation circuit can compensate the impedance of the car audio bus in real time by adjusting its own variable capacitor and inductor array based on a preset compensation strategy.

6. The method according to claim 3, characterized in that, The method further includes: The comparison analysis results, injection parameters, and corresponding bus status information are stored in the historical database, and the neural network diagnostic model is retrained based on the updated historical data to optimize the neural network diagnostic model.

7. The method according to claim 1, characterized in that, The digital chaotic sequence is generated based on a Logistic map or a Lorentz system model, and before being injected into the automotive audio bus, the digital chaotic sequence is modulated so that its spectrum continuously covers a frequency band from 1 MHz to 28 MHz.

8. An online diagnostic device for an automotive audio bus, characterized in that, Mounted on a host computer, the device includes: The sequence generation module is used to generate a digital chaotic sequence that is different from the current transmitted signal in the car audio bus; A signal injection module is used to inject the digital chaotic sequence into the car audio bus during the idle time slots when the transmission signal is periodically transmitted on the car audio bus, so as to send it to the target sub-node. A signal receiving module is used to receive a response signal transmitted back by an analog switch connected to the target sub-node through the car audio bus; wherein the analog switch is configured to transmit the signal forwarded from the sub-node back to the host, and the target sub-node forwards the digital chaotic sequence to the analog switch after determining that it cannot recognize the digital chaotic sequence. The signal comparison module is used to compare and analyze the received response signal as the target chaotic sequence with the digital chaotic sequence. The problem identification module is used to determine, based on the comparison analysis results, whether there is an impedance mismatch problem on the communication path from the host to the child node and the analog switch.

9. An electronic device, characterized in that, include: The device includes a processor, a storage medium, and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is in operation, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the online diagnostic method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the online diagnostic method as described in any one of claims 1 to 7.

Images