An ultrasonic ranging method and device based on a multi-parameter coupled compensation sensing network

Abstract

An ultrasonic ranging calibration method and apparatus based on a multi-parameter coupled compensation sensing network is disclosed. The method includes: first, constructing an ultrasonic ranging system to accurately acquire and transmit basic time parameters and multi-dimensional perturbation parameters; second, constructing a multi-parameter coupled compensation sensing network for distance calibration, fusing multiple parameters through a nonlinear perturbation injection function to construct an input vector, and using a hierarchical coupling compensation method to achieve step-by-step error correction, outputting a compensation amount or compensation factor based on the basic ranging result to complete distance correction; subsequently, collecting sample data under multiple perturbation parameter combinations and multiple reflection conditions, preprocessing the data, dividing it into training and test sets, and using an error constraint strategy based on the predicted deviation distribution for closed-loop training optimization to improve the model's generalization ability; finally, deploying the trained model to the main control processing unit to achieve real-time dynamic compensation ranging. This invention effectively overcomes the accuracy limitations of traditional ultrasonic ranging technology in complex environmental perturbations and multi-reflection characteristic adaptation scenarios, significantly improving ranging accuracy and scenario adaptability, and promoting the application expansion of ultrasonic ranging technology in complex scenarios such as industry, intelligent robots, and autonomous driving.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic ranging technology, and in particular to an ultrasonic ranging calibration method and device that dynamically compensates for basic ranging results by modeling multi-source disturbance parameters through coupling. Specifically, it is an ultrasonic ranging method and device based on a multi-parameter coupled compensation network (MPCCN). Background Technology

[0002] Ultrasonic ranging is a non-contact measurement technology that estimates distance based on the propagation delay of sound waves. Due to its simple structure, low cost, and ease of engineering implementation, it is widely used in fields such as industrial automation, intelligent robots, vehicle assistance systems, and complex environment detection.

[0003] In practical applications, ultrasonic ranging typically obtains a basic ranging result by measuring the time parameter of the sound wave from transmission to reception and combining it with a preset propagation model. However, this basic ranging result often contains various systematic biases, which originate not only from changes in the state of the acoustic propagation medium, but also from differences in the characteristics of the echo reflection interface, fluctuations in the system's operating state, and time stability.

[0004] From an engineering perspective, ultrasonic waves are affected by multiple disturbances during propagation and reflection. These disturbances exhibit highly coupled, nonlinear, and time-varying characteristics in different application scenarios. Traditional ranging methods typically correct the ranging results based on a single or limited number of parameters, or perform linear compensation through empirical models. These methods struggle to accurately characterize the complex relationships between multiple disturbances, leading to a significant decrease in ranging accuracy in complex environments, extreme conditions, or under multiple reflection conditions.

[0005] On the other hand, differences in the reflective interface have a significant impact on the shape of the echo signal. Under different reflection conditions, the echo signal exhibits significant differences in the time domain, frequency domain, and energy distribution. Existing ranging methods typically assume that the reflective interface conditions are constant or idealized, without systematically modeling reflection-related disturbances. As a result, they are prone to ranging instability or error amplification when facing non-ideal reflection scenarios.

[0006] In recent years, although some studies have attempted to introduce neural networks or data-driven models for ranging error correction, most solutions still remain at the stage of directly mapping basic time parameters to distance results. They lack structured modeling of the coupling relationships between multiple sources of disturbance, thus limiting the model's generalization ability and engineering adaptability. Furthermore, some models have complex structures and large parameter scales, making them unsuitable for real-time deployment in embedded ranging systems.

[0007] Therefore, there is an urgent need for a new ultrasonic ranging technology that can uniformly compensate for the systematic deviations introduced by multi-source disturbance factors in the basic ranging results without relying on explicit physical modeling, and has good real-time performance and engineering feasibility. Summary of the Invention

[0008] This invention aims to overcome the problems of unstable ranging accuracy, insufficient adaptability to reflection conditions, and limited system generalization ability of existing ultrasonic ranging technologies under complex disturbance environments. It provides an ultrasonic ranging method and device based on a multi-parameter coupled compensation sensing network, which achieves high-precision and high-stability ranging output by performing multi-dimensional disturbance coupling compensation on the basic ranging results.

[0009] The first aspect of this invention provides an ultrasonic ranging method based on a multi-parameter coupled compensation sensing network, the specific process of which is as follows: S1. Construct an ultrasonic ranging hardware system to achieve accurate acquisition and transmission of basic time parameters and multi-dimensional disturbance parameters (including acoustic propagation medium state, echo reflection characteristics, system working state and time stability related parameters); S2. Construct a multi-parameter coupled compensation sensing network, coupling and fusing the basic time parameters with the multi-dimensional perturbation parameter set to form a network input vector, and generating compensation information for correcting the basic ranging results through network inference; the network input vector is constructed through a perturbation coupling mapping function. S3. Collect sample data under various disturbance conditions and various reflection states, train the multi-parameter coupled compensation sensing network, and enable the network to learn the systematic deviation mapping relationship between the basic ranging results and the true distance. S4. Deploy the trained multi-parameter coupled compensation sensing network in the main control processing unit, collect input parameters in real time during the ranging process and perform model inference, and output the target distance value after dynamic compensation.

[0010] The hardware system described in step S1 includes an ultrasonic transducer unit, an ultrasonic signal processing unit, a multi-source parameter acquisition unit, and a main control processing unit. Specifically: the ultrasonic transducer unit transmits and receives ultrasonic signals; the ultrasonic signal processing unit extracts features from the echo signals to obtain fundamental time parameters; the multi-source parameter acquisition unit acquires a set of multi-dimensional perturbation parameters affecting ultrasonic propagation and reflection characteristics; and the main control processing unit performs ranging compensation calculations and model inference.

[0011] The multidimensional disturbance parameter set mentioned in step S1 includes one or more of the following parameter subsets: a first parameter subset used to characterize changes in the state of the acoustic propagation medium; a second parameter subset used to characterize differences in the reflection characteristics of echo signals; and a third parameter subset used to characterize the system's operating state and time stability.

[0012] The second parameter subset consists of multi-scale features of the echo signal in the time domain, frequency domain, and energy domain. These multi-scale features are fused to form a feature vector that characterizes the differences in the reflection interface and participates in the distance compensation process as part of the network input.

[0013] In step S2, the network input vector is constructed using a nonlinear perturbation injection function, the expression of which is:

[0014] in, Indicates the base time parameter. Represents a set of multidimensional perturbation parameters. This represents the perturbation feature mapping function. This represents the perturbation weighting coefficient. This indicates a weighted coupling operation.

[0015] Furthermore, the MPCCN network described in step S2 is a multi-layer feedforward structure, in which the internal layers achieve step-by-step error correction through a hierarchical coupling compensation method, and the mapping relationship of its k-th layer satisfies:

[0016] in, For network input vectors, This represents the basic mapping operator of the k-th layer. This represents the perturbation coupling compensation operator. This represents the incremental mapping of the perturbation parameter set to the current layer. This indicates a coupling operation.

[0017] Furthermore, the MPCCN network described in step S2 does not directly perform physical modeling and calculation of the target distance. Instead, it uses the baseline ranging results as a reference and corrects the baseline ranging results by outputting a compensation amount or compensation factor. The distance correction relationship satisfies:

[0018] in, The calibrated target distance value. Based on the distance measurement results, This is the compensation factor output by the network. This is a higher-order disturbance compensation correction function.

[0019] In step S3, the training process of the MPCCN network adopts an error constraint strategy based on the prediction bias distribution. By evaluating the deviation between the network output and the reference distance, when the compensation error meets the preset accuracy threshold, the training is determined to be complete and the model parameters are saved.

[0020] In a second aspect, the present invention also provides an ultrasonic ranging device based on a multi-parameter coupled compensation sensing network, comprising: an ultrasonic transducer unit for transmitting and receiving ultrasonic signals; an ultrasonic signal processing unit for extracting features from the echo signals and generating basic time parameters; a multi-source parameter acquisition unit for acquiring a set of multi-dimensional disturbance parameters that affect ranging accuracy; and a storage unit for storing the parameters of the trained multi-parameter coupled compensation sensing network model; wherein the main control processing unit is configured to execute the above method steps to dynamically correct the basic ranging results through the multi-parameter coupled compensation sensing network.

[0021] The main control processing unit is an embedded processor with model inference capability, and the storage unit includes a non-volatile memory for storing model parameters and a volatile memory for caching runtime data.

[0022] The main control processing unit is an embedded processor with model inference capability, and the storage unit includes a non-volatile memory for storing model parameters and a volatile memory for caching runtime data.

[0023] After obtaining the basic ranging results, the device dynamically corrects the basic ranging results through the multi-parameter coupled compensation sensing network, thereby maintaining the consistency and stability of the ranging results under different disturbance conditions.

[0024] Thirdly, the present invention also provides a computer-readable storage medium having program instructions stored thereon, which, when executed by a processor, implement the ultrasonic ranging method based on a multi-parameter coupled compensation sensing network of the present invention.

[0025] The working principle of this invention is: This invention uses baseline ranging results as a reference and does not explicitly model the ultrasonic propagation process. Instead, it constructs a multi-parameter coupled compensation sensing network to learn and compensate for systematic deviations introduced by multi-source disturbances. The compensation information output by the network is applied to the baseline ranging results to obtain calibrated target distance values, ensuring consistency and stability of the ranging results under different disturbance conditions.

[0026] The present invention has the following beneficial effects: 1. Clear compensation mechanism: The basic ranging results are corrected by outputting compensation information, the compensation location is clearly defined, and the physical ranging model is not directly replaced. 2. Multi-disturbance joint modeling: By handling multi-source disturbances through multi-parameter coupling, ranging stability in complex environments is significantly improved; 3. Strong generalization ability: The network learns systematic biases rather than single environmental characteristics, adapting to a variety of application scenarios; 4. Good engineering adaptability: The network structure can be customized and the parameter scale can be controlled, making it suitable for real-time deployment in embedded systems. Attached Figure Description

[0027] Figure 1 This is a flowchart of the method of the present invention.

[0028] Figure 2 This is a schematic diagram of the hardware system structure of the method of the present invention.

[0029] Figure 3 This is a diagram of the multi-parameter coupling compensation sensing network model architecture of the method of this invention.

[0030] Figure 4 This is a flowchart of the network model deployment process of the method of the present invention.

[0031] Figure 5 This is a schematic diagram of the device of the present invention. Detailed Implementation

[0032] To better understand the above technical solution, the technical solution of the ultrasonic ranging method and device based on a multi-parameter coupled compensation sensing network of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0033] Example 1

[0034] like Figures 1-5 As shown, this embodiment relates to an ultrasonic ranging method based on a multi-parameter coupled compensation sensing network, used to achieve high-precision ranging in conventional environments. The implementation process specifically includes the following steps: S1. Construct a multi-parameter sensing ultrasonic ranging hardware system to achieve accurate acquisition and transmission of basic time parameters and multi-dimensional disturbance parameters (including acoustic propagation medium state, echo reflection characteristics, system operating state, and time stability-related parameters); including the following modules: an ultrasonic transducer unit for transmitting and receiving ultrasonic signals; an ultrasonic signal processing unit for conditioning and feature extraction of echo signals to generate basic time parameters τ; a multi-source parameter acquisition unit for acquiring a set of multi-dimensional disturbance parameters P that affect ranging accuracy, including environmental parameters (temperature, humidity, air pressure), multi-scale characteristics of echo signals, system state, and time stability; a main control processing unit responsible for data preprocessing, model inference, and result output; and auxiliary modules providing stable power supply and interface output to achieve hardware collaborative operation.

[0035] S2. Design a multi-parameter coupled compensation sensing network (MPCCN) whose input vector consists of the fused basic time parameters and multi-dimensional perturbation parameters:

[0036] The network is a multi-layer feedforward structure, with each layer achieving step-by-step error correction through hierarchical coupling compensation:

[0037] in, Based on the mapping operator, For perturbation coupling compensation operator, This represents the disturbance increment.

[0038] S21. Sample Collection and Preprocessing: Sample data, including input feature vectors and true distance labels, are collected under various conventional environmental conditions and at various distances. First, outliers are removed using the 3σ criterion. Then, the min-max method is used to normalize the input features to the [0,1] interval. Finally, the training set and test set are divided in a 7:3 ratio.

[0039] S22. Network Training: During network training, the loss function uses mean squared error (MSE) to measure the deviation between the predicted and actual values; the optimizer uses Adam to update weights and biases; Dropout regularization is introduced in the hidden layers to prevent overfitting, with a dropout probability set to 0.2. Training iterates until the test set error meets the preset accuracy requirements, and the network model parameters are saved.

[0040] The MPCCN network described in step S2 does not directly perform physical modeling and calculation of the target distance. Instead, it uses the baseline ranging results as a reference and corrects the baseline ranging results by outputting a compensation amount or compensation factor. The distance correction relationship satisfies:

[0041] in, The calibrated target distance value. Based on the distance measurement results, This is the compensation factor output by the network. This is a higher-order disturbance compensation correction function.

[0042] S3. Model Deployment and Real-time Ranging: The trained MPCCN model is lightweighted and deployed to the main control processing unit to realize real-time acquisition of input parameters and inference, and output the dynamically compensated target distance d, so that the ranging results maintain consistency and stability under different disturbance conditions.

[0043] In step S3, the training process of the MPCCN network adopts an error constraint strategy based on the prediction bias distribution. By evaluating the deviation between the network output and the reference distance, when the compensation error meets the preset accuracy threshold, the training is determined to be complete and the model parameters are saved.

[0044] This invention first establishes an ultrasonic ranging system to accurately acquire and transmit basic time parameters and multi-dimensional perturbation parameters. Second, it constructs a multi-parameter coupled compensation sensing network for distance calibration. This network fuses multiple parameters through a nonlinear perturbation injection function to construct an input vector, and employs a hierarchical coupled compensation method to achieve step-by-step error correction. The compensation amount or factor is output as a reference based on the basic ranging result to complete the distance correction. Subsequently, sample data under multiple perturbation parameter combinations and multiple reflection conditions are collected. After preprocessing, the data is divided into training and testing sets. A closed-loop training optimization strategy based on the prediction deviation distribution is used to improve the model's generalization ability. Finally, the trained model is deployed to the main control processing unit to achieve real-time dynamic compensation ranging. This invention effectively overcomes the accuracy limitations of traditional ultrasonic ranging technology in complex environmental perturbations and multi-reflection characteristic adaptation scenarios, significantly improving ranging accuracy and scenario adaptability, and promoting the application expansion of ultrasonic ranging technology in complex scenarios such as industry, intelligent robots, and autonomous driving.

[0045] Example 2

[0046] This embodiment relates to an ultrasonic ranging method based on multi-parameter coupled compensation sensing network calibration. Based on Embodiment 1, to adapt to extremely short environments, the following steps are further taken: S1. Hardware system extreme environment adaptation optimization: Based on the hardware system architecture of Example 1, the core modules are optimized for extreme environment protection. Protective structures and materials adapted to extreme environments are adopted to ensure that core functions such as signal transmission and reception, parameter acquisition and signal processing are stably realized in extreme environments.

[0047] S2, Extreme Environment Adaptation Model Training.

[0048] S21. Under simulated extreme environmental conditions, sample data corresponding to different ranging distances are collected. The sample data includes input feature parameters and real distance labels under extreme environmental conditions. After preprocessing, the training set and test set are divided.

[0049] S22. Adaptive training of the multi-parameter coupled compensation sensing network based on the training set under extreme environments, optimizing the training strategy based on Example 1: 1) A learning rate decay strategy is adopted, which reduces the learning rate as the number of iterations increases to adapt to the special characteristics of sample distribution in extreme environments. The learning rate decay formula is:

[0050] in, for Learning rate at any given moment The initial learning rate, The attenuation coefficient is... This is the decay step size; 2) Add environmental stress factor weights to the input features to enhance the model's adaptability to extreme environmental parameter changes. Train the model to the MAE test set to meet the accuracy and stability requirements under extreme environments.

[0051] S3. Deployment and Extreme Environment Testing: The model adapted for extreme environments was deployed to the optimized hardware system, and ranging tests were conducted under extreme conditions. Test results show that this method can achieve stable ranging under extreme conditions, meeting the usage requirements of extreme scenarios.

[0052] Example 3

[0053] This embodiment relates to an ultrasonic ranging method based on multi-parameter coupled compensation sensing network calibration, used to achieve real-time ranging in multi-material mixed scenes. The implementation process is based on Embodiment 1, with the core addition of material adaptation-related design, specifically including the following steps: S1. Hardware system expansion and adaptation: Based on the hardware system of Example 1, a material switching mechanism is added to realize the orderly switching of different reflective materials and ensure the simulation and testing of multi-material mixed scenes; the multi-parameter sensing module is optimized to improve the acquisition accuracy of the reflection characteristics of different materials.

[0054] S2, Multi-material Adaptation Model Training

[0055] S21. In the target indoor environment, different reflective materials are switched through the material switching mechanism, and sample data corresponding to different distances of each material are collected, including input feature parameters and real distance labels. After preprocessing, the data is divided into training set and test set.

[0056] S22. Train a multi-parameter coupled compensation perception network based on a multi-material sample training set to optimize the model's feature learning ability: 1) A material feature encoding module is added to the model input layer to map different material types into one-hot encoded vectors, which are then concatenated with the original feature vectors and input into the network to enhance the distinguishability of material features; 2) The loss function incorporates a material adaptation penalty term, and the formula is optimized as follows:

[0057] in, This is the original mean square error loss. The penalty coefficient is... The number of samples for different materials. For the first Predicted values ​​for each material sample. For the first The true labels of each material sample are used to ensure consistent model adaptation across all materials through this optimization. 3) Employ an early stopping strategy: stop training when the validation set error fails to decrease for multiple consecutive rounds to avoid overfitting. Train until the model meets the accuracy and real-time response requirements for various material scenarios.

[0058] S3. Deployment and Dynamic Switching Test: The multi-material adaptation model was deployed to the expanded hardware system, and the material switching mechanism was activated to achieve dynamic switching between different materials. Real-time ranging tests were conducted. Test results show that this method can achieve stable and real-time ranging in multi-material mixed scenarios and has good material adaptability.

[0059] Example 4

[0060] This embodiment relates to an ultrasonic ranging device based on multi-parameter coupled compensation sensing network calibration, such as... Figure 5 It includes a memory and one or more processors. The memory stores executable code. When the one or more processors execute the executable code, they are used to implement the ultrasonic ranging method based on multi-parameter coupling compensation sensing network calibration described in any of the above embodiments 1 to 3.

[0061] At the hardware level, the device includes the functional modules described in the foregoing embodiments, and each module interacts and works collaboratively through a preset interface. The processor reads the corresponding computer program from the memory and runs it to implement the above-described ranging method. Besides software implementation, this invention does not exclude other implementation methods, such as logic devices or a combination of hardware and software.

[0062] Example 5

[0063] This embodiment relates to a computer-readable storage medium storing a program that, when executed by a processor, implements the ultrasonic ranging method based on multi-parameter coupling compensation sensing network calibration as described in any of embodiments 1 to 3 above.

[0064] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device for calibrating ultrasonic ranging functions using a multi-parameter coupled compensation sensing network specified in one or more boxes.

[0065] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function of calibrating ultrasonic ranging using a multi-parameter coupled compensation sensing network specified in one or more boxes.

[0066] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps for calibrating ultrasonic ranging functionality using a multi-parameter coupled compensation sensing network specified in one or more boxes.

[0067] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0068] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0069] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0070] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0071] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0072] This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0073] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0074] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. An ultrasonic ranging calibration method based on a multi-parameter coupled compensation sensing network, characterized in that, Includes the following steps: S1. Construct an ultrasonic ranging hardware system to achieve accurate acquisition and transmission of basic time parameters and multidimensional disturbance parameters; S2. Construct a multi-parameter coupled compensation sensing network (MPCCN) for distance calibration, fuse the basic time parameters with the multi-dimensional perturbation parameter set to form a network input vector, and output the target distance value after nonlinear compensation. The network input vector is constructed through a nonlinear perturbation injection function. S3. Collect sample data under various combinations of perturbation parameters and various reflection conditions, train the MPCCN network, and enable it to learn the systematic deviation relationship between the basic time parameters and the true distance due to environmental perturbation, medium change and reflection characteristics differences, and form a distance compensation mapping model. S4. Deploy the trained MPCCN to the main control processing unit, collect input parameters in real time during the ranging process and perform model inference, and output the target distance value after dynamic compensation to weaken or eliminate the influence of multi-source disturbance factors on the accuracy of ultrasonic ranging.

2. The method according to claim 1, characterized in that, The hardware system described in step S1 includes an ultrasonic transducer unit, an ultrasonic signal processing unit, a multi-source parameter acquisition unit, and a main control processing unit, wherein: the ultrasonic transducer unit is used to transmit and receive ultrasonic signals; the ultrasonic signal processing unit is used to extract features from the echo signals to obtain basic time parameters; the multi-source parameter acquisition unit is used to acquire a set of multi-dimensional disturbance parameters that affect the propagation and reflection characteristics of ultrasound; and the main control processing unit is used to perform ranging compensation calculations and model inference.

3. The method according to claim 2, characterized in that, The multidimensional perturbation parameter set mentioned in step S1 includes one or more of the following parameter subsets: a first parameter subset used to characterize the state changes of the acoustic propagation medium; A second subset of parameters used to characterize the differences in the reflection characteristics of echo signals; The third subset of parameters used to characterize the system's operating state and time stability.

4. The method according to claim 1, characterized in that, In step S2, the network input vector is constructed using a nonlinear perturbation injection function, the expression of which is: in, Indicates the base time parameter. Represents a set of multidimensional perturbation parameters. This represents the perturbation feature mapping function. This represents the perturbation weighting coefficient. This indicates a weighted coupling operation.

5. The method according to claim 1, characterized in that, The MPCCN network described in step S2 is a multi-layer feedforward structure. Its internal layers achieve step-by-step error correction through a hierarchical coupling compensation method. The mapping relationship of its k-th layer satisfies: in, For network input vectors, This represents the basic mapping operator of the k-th layer. This represents the perturbation coupling compensation operator. This represents the incremental mapping of the perturbation parameter set to the current layer. This indicates a coupling operation.

6. The method according to claim 4, characterized in that, The MPCCN network described in step S2 uses the base ranging result as a reference and corrects the base ranging result by outputting a compensation amount or compensation factor. The distance correction relationship satisfies: in, The calibrated target distance value. Based on the distance measurement results, This is the compensation factor output by the network. This is a higher-order disturbance compensation correction function.

7. The method according to claim 3, characterized in that, The second parameter subset consists of multi-scale features of the echo signal in the time domain, frequency domain, and energy domain. These multi-scale features are fused to form a feature vector that characterizes the differences in the reflection interface and participates in the distance compensation process as part of the network input.

8. The method according to claim 1, characterized in that, In step S3, the training process of the MPCCN network adopts an error constraint strategy based on the prediction bias distribution. By evaluating the deviation between the network output and the reference distance, when the compensation error meets the preset accuracy threshold, the training is determined to be complete and the model parameters are saved.

9. An ultrasonic ranging device based on a multi-parameter coupled compensation sensing network, characterized in that, include: The ultrasonic transducer unit is used to transmit and receive ultrasonic signals; the ultrasonic signal processing unit is used to extract features from the echo signals and generate basic time parameters. A multi-source parameter acquisition unit is used to acquire a set of multi-dimensional disturbance parameters that affect ranging accuracy; a storage unit is used to store the parameters of the trained multi-parameter coupled compensation sensing network model; and a main control processing unit is used to execute the method of any one of claims 1 to 8 and output the compensated target distance value through model inference.

10. The apparatus according to claim 9, characterized in that, The main control processing unit is an embedded processor with model inference capability. The storage unit includes a non-volatile memory for storing model parameters and a volatile memory for caching runtime data. After obtaining the basic ranging results, the device dynamically corrects the basic ranging results through the multi-parameter coupled compensation sensing network, thereby maintaining the consistency and stability of the ranging results under different disturbance conditions.

Images