Indoor person tracking method based on electrostatic sensing technology

Abstract

The invention discloses an indoor personnel tracking method based on an electrostatic sensing technology, and the method comprises the steps: obtaining electrostatic signals when a person reaches different positions of a room through an electrostatic sensor array, a signal conditioning circuit and a microprocessor unit; training a recurrent neural network according to the electrostatic signal time sequence and the personnel position, and deploying a trained recurrent neural network model to a microprocessor; according to electrostatic signals generated by indoor personnel activities, an electrostatic signal time sequence is acquired in real time; and inputting an electrostatic signal time sequence acquired in real time into the microprocessor, and predicting the position of the indoor personnel through the trained recurrent neural network model to obtain a position prediction result. According to the invention, people do not need to carry any equipment, and real-time tracking of the position of the people can be realized through the electrostatic sensor array.

Description

Technical Field

[0001] The present invention belongs to the technical field of indoor personnel tracking, and in particular relates to an indoor personnel tracking method based on electrostatic sensing technology. Background Art

[0002] With the increasing demand for smart security systems and surveillance in specialized locations, there is an urgent need for non-invasive personnel location and tracking technology in nursing homes, educational institutions, and judicial oversight. In elderly care settings, traditional infrared positioning systems cannot effectively distinguish between normal daily activities and the initial position changes of elderly individuals during sudden fainting. The UWB wristband solutions widely used in educational institutions carry the risk of accidental removal by children and are unable to detect detailed behavioral trajectories. RFID area positioning technology used in factories and other places requires personnel to carry RFID tags. Existing monitoring equipment suffers from high maintenance costs, poor compliance, and the requirement for personnel to carry part of the equipment. Video surveillance solutions also face fundamental limitations in terms of privacy protection and adaptability to complex environments.

[0003] Non-contact monitoring technology based on electrostatic induction offers a new solution for these scenarios. The changes in the spatial electric field distribution caused by human motion can be analyzed through electrode array signals to determine positional changes. These signal characteristics provide a physical basis for modeling human motion. For example, low-frequency displacement components correspond to normal activity, while high-frequency components are associated with the details of human movement. Compared to traditional optical and radio frequency solutions, this technology offers the unique advantages of strong physical concealment, high environmental compatibility, and the absence of biometric data collection, demonstrating significant application value in sensitive and controlled environments.

[0004] Existing electrostatic sensing technology faces multiple technical bottlenecks in indoor people tracking applications. In elderly care settings, the frequent movement of caregivers couples with the motion signals of monitored targets, leading to high false alarm rates in traditional threshold detection methods. In educational settings, the concurrent movement of multiple targets causes significant changes in the electrostatic field, making it impossible for existing single-electrode systems to decouple person location from motion trajectory. These issues hinder the large-scale application of electrostatic sensing technology in people tracking. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an indoor personnel tracking method based on electrostatic sensing technology to meet the tracking needs of key personnel in various scenarios such as nursing homes and kindergartens.

[0006] To achieve the above object, the present invention provides an indoor personnel tracking method based on electrostatic sensing technology, comprising:

[0007] Through the electrostatic sensor array, signal conditioning circuit and microprocessor unit, the electrostatic signals of people reaching different positions in the room are obtained;

[0008] Based on the electrostatic signal time series and the personnel location, a recurrent neural network is trained and deployed on a microprocessor.

[0009] According to the static electricity signals generated by indoor human activities, the time series of static electricity signals is acquired in real time;

[0010] The real-time collected electrostatic signal time series is input into the microprocessor, and the position of indoor personnel is predicted through the trained recurrent neural network model to obtain the position prediction result.

[0011] Preferably, the process of obtaining electrostatic signals of people reaching different positions in the room by using an electrostatic sensor array, a signal conditioning circuit, and a microprocessor unit includes:

[0012] The induced charge signal received by the electrostatic sensor array is obtained through a signal conditioning circuit and a microprocessor.

[0013] Preferably, the induced charge signal is expressed as:

[0014]

[0015] Where, represents the surface of the electrostatic sensor electrode, ε0 is the dielectric constant in vacuum, E(r,t) is the electric field strength caused by the charge in the measurement space, r and t represent position and time respectively, and n is the normal vector of the electrode surface.

[0016] Preferably, the electric field intensity E(r, t) caused by the charge in the measurement space is expressed as:

[0017]

[0018] Where Ω represents the measurement space, q(t) is the charge, and r q is the location of the charge.

[0019] Preferably, the process of obtaining the induced charge signal received by the electrostatic sensor array through the signal conditioning circuit and the microprocessor includes:

[0020] Install electrostatic sensors at detection positions on the walls and ceiling of the room to form an electrostatic sensor array;

[0021] The signal of each electrostatic sensor is amplified and filtered by the signal conditioning circuit, and the electrostatic analog signal is received by an independent microprocessor unit and converted into a digital signal by the built-in ADC to obtain the induced charge signal.

[0022] Preferably, the process of obtaining electrostatic signals of people reaching different positions in the room by using an electrostatic sensor array, a signal conditioning circuit, and a microprocessor unit includes:

[0023] Connect the power supply to the power conversion circuit to realize the power supply of the electrostatic sensor array, signal conditioning circuit and microprocessor;

[0024] The electrostatic sensor array and the microprocessor unit are used to obtain the electrostatic sensor signals installed at different locations on the walls and ceiling of the room;

[0025] The electrostatic sensor signals when people are at different positions in the room are obtained through the signal conditioning circuit. Based on the independent microprocessor unit, the analog signal of each electrostatic sensor is converted into a digital signal through the built-in ADC to obtain the induced charge signal.

[0026] Preferably, the process of training a recurrent neural network based on the electrostatic signal time series and the personnel position and deploying the trained recurrent neural network model on a microprocessor includes:

[0027] The static electricity signals of indoor people at different positions, motion postures, and environmental conditions are obtained. The person's location is obtained based on the distance between the UWB tag carried by the person and the UWB anchor points arranged in the room to form a training sample set.

[0028] A recurrent neural network model is trained according to the training sample set, wherein the input of the trained recurrent neural network model is the time series signal of the electrostatic sensor array, and the output is the position of the personnel.

[0029] Compared with the prior art, the present invention has the following advantages and technical effects:

[0030] The solution presented in this paper addresses the problem of indoor person tracking under diverse environmental and activity conditions. It uses an electrostatic sensor array to capture the electrostatic signals generated by human activity and, using a recurrent neural network deployed within a microprocessor, predicts the person's location. This allows for real-time tracking of a person's location without the need for the person to carry any equipment. By processing the electrostatic signals using a recurrent neural network model, it effectively decouples the concurrent motion signals of multiple targets, improving positioning accuracy.

[0031] This invention uses electrostatic sensing technology, which has strong physical concealment, high environmental compatibility, and does not involve biometric feature collection, making it suitable for a variety of complex environments. It solves the problem of high false alarm rate of traditional threshold detection methods in complex scenarios and improves system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0033] Figure 1 This is a schematic diagram of indoor personnel position tracking according to an embodiment of the present invention;

[0034] Figure 2 is a structural diagram of an electrostatic sensing system according to an embodiment of the present invention;

[0035] Figure 3 This is a structural diagram of a recurrent neural network model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0037] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0038] Example 1

[0039] like Figure 1-3 As shown, this embodiment provides an indoor person tracking method based on electrostatic sensing technology, including:

[0040] Through the electrostatic sensor array, signal conditioning circuit and microprocessor unit, the electrostatic signals of people reaching different positions in the room are obtained;

[0041] Based on the electrostatic signal time series and the personnel location, a recurrent neural network is trained and deployed on a microprocessor.

[0042] According to the static electricity signals generated by indoor human activities, the time series of static electricity signals is acquired in real time;

[0043] The real-time collected electrostatic signal time series is input into the microprocessor, and the position of indoor personnel is predicted through the trained recurrent neural network model to obtain the position prediction result.

[0044] This embodiment uses the coordinated operation of an electrostatic sensor array, a signal conditioning circuit, and a microprocessor unit to accurately capture the electrostatic signals generated by indoor human activity. The signal conditioning circuit amplifies and filters the induced charge signals received by the electrostatic sensors, effectively removing noise interference and improving signal quality, providing a reliable data foundation for subsequent accurate location prediction.

[0045] Furthermore, the solution specifically includes the following steps: Step 1: obtaining electrostatic signals of people reaching different positions in the room through an electrostatic sensor array, a signal conditioning circuit, and a microprocessor unit; specifically, obtaining an induced charge signal received by the electrostatic sensor array through the signal conditioning circuit and the microprocessor;

[0046] In a preferred embodiment, the induced charge signal generated on the electrode due to the activity of people in the room in step 1 is expressed as:

[0047]

[0048] Where, represents the surface of the electrostatic sensor electrode, ε0 is the dielectric constant in vacuum, E(r,t) is the electric field strength caused by the charge in the measurement space, r and t represent position and time respectively, and n is the normal vector of the electrode surface.

[0049] The electric field strength can be calculated by the following formula:

[0050]

[0051] Where Ω represents the measurement space, q(t) is the charge, and r q is the location of the charge.

[0052] In a preferred embodiment, step 1 involves installing electrostatic sensors at multiple detection locations on the walls and ceiling of a room to form an electrostatic sensor array. The signal from each electrostatic sensor is amplified and filtered by a signal conditioning circuit. An independent microprocessor receives the electrostatic analog signal and converts it into a digital signal using a built-in ADC to obtain an induced charge signal. The signal conditioning circuit and microprocessor of the electrostatic sensor array are powered by a unified power supply.

[0053] Specifically, the electrostatic sensor array and microprocessor unit are used to acquire electrostatic signals from various locations on the room's walls and ceiling. First, a power supply is connected to a power conversion circuit to power the electrostatic sensor array, signal conditioning circuit, and microprocessor. The signal conditioning circuit then acquires electrostatic sensor signals from people at various locations within the room. An independent microprocessor unit then converts the analog signal from each electrostatic sensor into a digital signal using a built-in ADC.

[0054] This embodiment uses a non-contact monitoring method based on electrostatic sensing technology, eliminating the need for personnel to wear any equipment (such as UWB wristbands or RFID tags), thus avoiding the inconvenience and poor compliance caused by wearing equipment. In settings such as nursing homes and kindergartens, the elderly and children are often unwilling or unable to wear equipment for long periods of time. The non-invasive nature of this invention can significantly improve the system's acceptability and practicality.

[0055] This embodiment's electrostatic sensing technology doesn't involve biometric collection, offers strong physical concealment and environmental compatibility, and neither infringes on personal privacy nor disrupts the indoor environment. In locations where privacy protection is crucial (such as nursing homes and judicial oversight facilities), this embodiment can meet stringent privacy protection requirements while adapting to a variety of complex indoor environments.

[0056] Step 2: Based on the electrostatic signal time series and the personnel location, a recurrent neural network (RNN) is trained and deployed on the microprocessor.

[0057] In a preferred embodiment, step 2 captures electrostatic signals from people in different room locations, motion patterns, and environmental conditions. The person's location is determined by the distance between the person's UWB tag and UWB anchor points located in the room. This constitutes a training sample set used to train a recurrent neural network model. The model input is the time series signals from the electrostatic sensor array, and the output is the person's location. The trained model and network parameters are deployed in a microprocessor unit, which receives the electrostatic digital signals sent by other microprocessors.

[0058] Furthermore, the electrostatic signal is normalized using formula (3):

[0059]

[0060] Wherein, i represents the i-th electrostatic electrode, and n represents the number of electrostatic sensors.

[0061] At time t, the multi-frame electrostatic signal of the electrostatic sensor array is the model input (m is the number of signal sampling points in a frame), the personnel position coordinate x t ,y t The model output is a one-dimensional convolutional neural network (1D-CNN) model that extracts input signal features. A recurrent neural network is used to construct a regression model between the signal features and the person's location. The model loss function is the error between the predicted location and the actual location. During the model optimization process, model training is achieved by minimizing the loss function, and the trained neural network model is deployed on the microprocessor chip.

[0062] This embodiment uses a recurrent neural network (RNN) to process and analyze the time series of electrostatic signals, which can capture the movement trajectory and position information of people contained in the electrostatic signals. The RNN model has a powerful time series modeling capability and can handle complex signal dynamic changes, thereby achieving high-precision personnel position prediction. Compared with the traditional threshold detection method, the positioning accuracy of the present invention is greatly improved, which can effectively solve the problem of high false alarm rate in complex scenarios, especially in the elderly care scenario where the frequent movement of caregivers is coupled with the motion signals of the monitoring targets, and the target personnel can still be accurately distinguished and located.

[0063] Furthermore, in scenarios with multiple targets moving concurrently, such as educational venues, existing single-electrode systems cannot decouple individual positioning from their motion trajectories. However, this embodiment, through the combination of an electrostatic sensor array and a recurrent neural network model, effectively decouples the electrostatic signals from multiple targets, enabling independent positioning and motion tracking of each target, thus resolving the bottleneck of existing technologies in multi-target scenarios.

[0064] Step 3: In the application stage, based on the electrostatic signals generated by the activities of people indoors, the time series of electrostatic signals are collected in real time, and the location of people indoors is predicted through a microprocessor;

[0065] In a preferred embodiment, in step 3, under the condition of unknown indoor personnel location information, a recurrent neural network model is used to predict the personnel location at the current moment by weighted summation based on the location at the previous moment and the electrostatic signal obtained at the current moment.

[0066] In this embodiment, the electrostatic sensor is a circular electrode, but other geometric structures may also be used; personnel position prediction uses a recurrent neural network model, but other mechanism models or artificial intelligence methods may also be used; the neural network model is deployed on a microprocessor, but other processors may also be used.

[0067] The electrostatic sensor array, signal conditioning circuit, and microprocessor unit in this embodiment are simple in structure and easy to install and maintain. Compared with traditional infrared positioning systems, UWB wristband solutions, and RFID area positioning technologies, this embodiment has lower equipment costs and does not require frequent replacement or maintenance, reducing the overall maintenance cost of the system.

[0068] This embodiment uses a signal conditioning circuit to pre-process electrostatic signals, and a recurrent neural network model to deeply learn and analyze these signals. This effectively addresses signal interference and noise in various complex environments, improving system stability and reliability. Even in scenarios with frequent human activity and significant environmental fluctuations, the system maintains stable operation, providing reliable support for personnel tracking and monitoring.

[0069] In summary, this embodiment achieves high-precision and high-robustness indoor personnel positioning and tracking through the organic combination of electrostatic sensor array, signal conditioning circuit, microprocessor unit and recurrent neural network model, and at the same time has significant advantages such as non-invasiveness, high compliance, wide applicability, low maintenance cost and high reliability.

[0070] Example 2

[0071] This embodiment further discloses a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first embodiment.

[0072] Example 3

[0073] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first embodiment are implemented.

[0074] Example 4

[0075] This embodiment further discloses a computer program product, including a computer program, which implements the steps of the method described in the first embodiment when executed by a processor.

[0076] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for indoor personnel tracking based on electrostatic sensing technology, characterized in that: include: Through the electrostatic sensor array, signal conditioning circuit and microprocessor unit, the electrostatic signals of people reaching different positions in the room are obtained; Based on the electrostatic signal time series and the personnel location, a recurrent neural network is trained and deployed on a microprocessor. According to the static electricity signals generated by indoor human activities, the time series of static electricity signals is acquired in real time; The real-time collected electrostatic signal time series is input into the microprocessor, and the position of indoor personnel is predicted through the trained recurrent neural network model to obtain the position prediction result.

2. The method according to claim 1, characterized in that The process of obtaining electrostatic signals of people reaching different locations in the room through an electrostatic sensor array, a signal conditioning circuit, and a microprocessor unit includes: The induced charge signal received by the electrostatic sensor array is obtained through a signal conditioning circuit and a microprocessor.

3. The method according to claim 2, characterized in that The induced charge signal is expressed as: Where, represents the surface of the electrostatic sensor electrode, ε0 is the dielectric constant in vacuum, E(r,t) is the electric field strength caused by the charge in the measurement space, r and t represent position and time respectively, and n is the normal vector of the electrode surface.

4. The method according to claim 3, characterized in that The electric field intensity E(r,t) caused by the charge in the measurement space is expressed as: Where Ω represents the measurement space, q(t) is the charge, and r q is the location of the charge.

5. The method according to claim 2, characterized in that The process of obtaining the induced charge signal received by the electrostatic sensor array through the signal conditioning circuit and the microprocessor includes: Install electrostatic sensors at detection positions on the walls and ceiling of the room to form an electrostatic sensor array; The signal of each electrostatic sensor is amplified and filtered by the signal conditioning circuit, and the electrostatic analog signal is received by an independent microprocessor unit and converted into a digital signal by the built-in ADC to obtain the induced charge signal.

6. The method according to claim 1, characterized in that The process of obtaining electrostatic signals of people reaching different locations in the room through an electrostatic sensor array, a signal conditioning circuit, and a microprocessor unit includes: Connect the power supply to the power conversion circuit to realize the power supply of the electrostatic sensor array, signal conditioning circuit and microprocessor; The electrostatic sensor array and the microprocessor unit are used to obtain the electrostatic sensor signals installed at different locations on the walls and ceiling of the room; The electrostatic sensor signals when people are at different positions in the room are obtained through the signal conditioning circuit. Based on the independent microprocessor unit, the analog signal of each electrostatic sensor is converted into a digital signal through the built-in ADC to obtain the induced charge signal.

7. The method according to claim 1, characterized in that The process of training a recurrent neural network based on the electrostatic signal time series and the personnel location and deploying the trained recurrent neural network model on a microprocessor includes: The static electricity signals of indoor people at different positions, motion postures, and environmental conditions are obtained. The person's location is obtained based on the distance between the UWB tag carried by the person and the UWB anchor points arranged in the room to form a training sample set. A recurrent neural network model is trained according to the training sample set, wherein the input of the trained recurrent neural network model is the time series signal of the electrostatic sensor array, and the output is the position of the personnel.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images