An ultrasonic-based static obstacle detection system and method

Abstract

The application discloses an ultrasonic-based static obstacle detection system and method, which comprises a static obstacle detection device and a central server, wherein the static obstacle detection device is installed in a place where static obstacle detection is needed and is used for obstacle detection; the central server is used for receiving, storing and managing data sent by the static obstacle detection device and sending an alarm notification when a static obstacle exists; the application adopts a differential detection method and a static obstacle filtering algorithm, can effectively adapt to changes of various environments, is not sensitive to environmental factors such as pedestrians, air flow, temperature and light changes, and has low false alarm and missed alarm rates.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of wireless intelligent sensing of Internet of Things, and particularly relates to a system and method for sensing static obstacles based on ultrasonic waves. BACKGROUND

[0002] With the development of social economy, safety accidents caused by the blockage of fire escape passages by sundries are common, and fires caused by the charging of electric vehicles in building corridors occur frequently, resulting in great loss of people's personal and property safety. Therefore, object detection technology for security has attracted widespread attention and research, and the main purpose is to timely detect large static objects in fire passages or other occasions, have early warning and centralized management functions, and are easy to deploy on a large scale to prevent safety hazards.

[0003] The current mainstream obstacle detection methods are:

[0004] 1) traditional manual inspection, which is not only time-consuming and labor-intensive, but also cannot achieve timely detection and timely processing;

[0005] 2) video monitoring, which is a widely used solution with strong real-time performance and good effect, but in non-public places such as residential buildings, it is often difficult to deploy due to privacy issues;

[0006] 3) infrared sensing, which has the advantages of low cost and high reliability, but can only be deployed in the form of infrared rays and infrared light curtains, has a small coverage range, and has poor applicability in area monitoring;

[0007] 4) new sensing technologies such as millimeter wave radar or laser radar, which have wide detection range and high precision, but the cost is too high.

[0008] Therefore, based on the above considerations, it is necessary to propose an innovative obstacle sensing system that uses sensors to monitor in real time whether there are objects such as electric vehicles in the target area that cause fire hazards, improve the real-time performance, reliability and safety of fire monitoring, and avoid infringing on the privacy of residents. SUMMARY

[0009] In view of the shortcomings of the prior art, the purpose of the present application is to provide a static obstacle detection system based on ultrasonic waves and a working method thereof, to solve the problems of low precision, poor real-time performance, high cost and possible privacy infringement of existing obstacle detection technologies, and to detect and alarm newly added fixed objects in the target environment, while having the ability to exclude dynamic interference such as pedestrians and the environment.

[0010] To achieve the above purpose, the technical solution adopted by the present application is as follows:

[0011] The application discloses an ultrasonic-based static obstacle detection system, which comprises a static obstacle detection device and a central server.

[0012] The static obstacle detection device comprises a microcomputer, a microphone array and a speaker array.

[0013] The microcomputer is used for controlling the speaker array to emit ultrasonic waves and the microphone array to receive the reflected ultrasonic wave signals, and processing the received reflected signals to determine whether a static obstacle exists, and uploading the result data to the central server through a network.

[0014] Further, the speaker array has one or more speakers which are connected to a power amplifier module through a line and then connected to an audio socket of the microcomputer to emit ultrasonic waves with a specific waveform.

[0015] Further, the microphone array has one or more omnidirectional microphones which are connected to the microcomputer through a USB, a serial port or a network port to receive the reflected ultrasonic wave signals.

[0016] Further, the ultrasonic waves emitted by the speaker array are composed of frequency-modulated continuous waves (FMCW) with adjustable frequency range, time length and repetition times.

[0017] The application further discloses an ultrasonic-based static obstacle detection method based on the above system, which comprises the following steps.

[0018] (1) initialization: obtaining an initial detection result y0(d) through an environment detection algorithm in an open environment, wherein d is the distance between an obstacle and the device, and the initial detection result is a curve with distance as the horizontal coordinate and obstacle intensity as the vertical coordinate;

[0019] (2) real-time detection: continuously running the environment detection algorithm at a certain time interval, and obtaining a real-time detection result y k (d) each time;

[0020] (3) threshold setting: setting an adaptive detection threshold Y(d) = Y0 / d 2 , wherein Y0 is an empirical value and is (0, 1);

[0021] (4) differential analysis: making a difference between the real-time detection result y k (d) and the initial detection result y0(d), and if there is a distance d, y k (d) - y0(d) > 0, an alarm is sent.(d) -y0(d) > Y(d), that is, if the difference exceeds the adaptive detection threshold, it is determined that there is a new static obstacle at the distance d;

[0022] (5) The detected static obstacles are filtered according to the static obstacle filtering algorithm, and the static obstacles meeting the conditions are sent to the central server.

[0023] Further, the open environment refers to the target site to be detected in the initial case without obstacles, and all building structures and objects in the environment will be regarded as background and not detected as obstacles.

[0024] Further, the static obstacle refers to an object with a large volume that is distinguished from the background, newly added and stably present, and pedestrians, animals or other transient objects and interference are excluded.

[0025] Further, the environment detection algorithm specifically includes:

[0026] (a) generating a signal to be sent: the signal to be sent is composed of N frames, each frame contains a linear frequency modulation wave wav sig with a time length of t sig and t int seconds of blank, wherein the frequency of the linear frequency modulation wave wav sig is f min ~f max ;

[0027] (b) sending and receiving ultrasonic waves: the signal to be sent is played through a loudspeaker array, and the audio signal received by a microphone array is recorded as wav recv ;

[0028] (c) filtering noise: the received audio signal is filtered through a band-pass filter with a frequency range of f min ~f max , so as to filter out noise interference of other frequency spectrums;

[0029] (d) calculating correlation: calculating the normalized cross-correlation function ccf of the above filtered audio signal and the linear frequency modulation wave wav sig , and the value range of the normalized cross-correlation function is [0, 1];

[0030] (e) multi-cycle mean filtering: and the peak value detection of the normalized cross-correlation function ccf is up to N points, according to the peak points p1, p2,..., p n , the normalized cross-correlation function ccf is segmented and restored into N frames, in order to reduce the influence of Gaussian noise dynamic targets, the N frames are superimposed to obtain the mean value, and the multi-cycle mean filtering is performed;

[0031] (f) Extracting the upper envelope of the above mean filtered curve to eliminate the phase fluctuation of the waveform, and obtaining a stable environmental detection result y(d).

[0032] Further, the function expression of the linear frequency modulation wave is:

[0033]

[0034] Wherein, A is the amplitude of the signal, is the center frequency of the signal, B=F h -F l is the bandwidth of the sweep, F h and F l are the highest and lowest scanning frequencies, t is time, T is scanning time, and φ0 is initial phase.

[0035] Wherein, the calculation formula of the cross-correlation function is:

[0036]

[0037] Wherein, f1(t) and f2(t) are two function sequences with time as the independent variable, t is time, and τ is the integral variable; Normalization means scaling the data between 0 and 1 according to the maximum and minimum values, and the formula is: x max is the maximum value of the data, x min is the minimum value of the data.

[0038] Further, the static obstacle filtering algorithm filters the detected static obstacles, that is, the obstacles need to meet the following conditions to report to the central server, otherwise it is judged as a pedestrian or environmental interference:

[0039] (51) If the static obstacle detection device contains multiple microphones, a certain number of microphones detect obstacles (the parameter can be adjusted, and generally can be set to 1 / 2);

[0040] (52) The obstacle exists for a certain time (the parameter can be adjusted, and generally can be set to one minute).

[0041] The beneficial effects of the present application are:

[0042] 1. High-precision static obstacle detection: In a fixed environment, it can effectively detect the electric vehicles, waste paper boxes and other large-volume obstacles placed by people, avoid blocking the fire evacuation channel and cause fire hazards;

[0043] 2. High reliability and high robustness: due to the adoption of differential detection method and static obstacle filtering algorithm, the application can effectively adapt to changes in various environments, filter pedestrians and other dynamic targets, and is not sensitive to environmental factors such as air flow, temperature and light changes, with low false positive and false negative rates;

[0044] 3. Wide coverage: compared with traditional ultrasonic probes, the application adopts adaptive detection threshold, and the effective detection angle is about 90° in front of the device, and the maximum detection distance is about 5 meters;

[0045] 4. Low device complexity and cost: the system uses common microphones, speakers, power amplifier boards and other devices, which are lower in cost than similar technologies such as millimeter wave and laser radar, and can be deployed on a large scale;

[0046] 5. Strong scalability and easy large-scale deployment: the system uses common microphones, speakers, microcomputers and other general-purpose hardware, which can not only be used for obstacle detection, but also can add positioning, broadcasting, communication and other Internet of Things functions, and can be deployed on a large scale in scenarios without network access after using GPRS or 4G antennas. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 The system architecture diagram of the application.

[0048] Figure 2 The static obstacle detection device schematic diagram of the application.

[0049] Figure 3 The static obstacle detection overall flowchart of the application.

[0050] Figure 4 The environment detection algorithm flowchart of the application.

[0051] Figure 5 The FMCW ultrasonic signal schematic diagram of the application. DETAILED DESCRIPTION

[0052] In order to facilitate the understanding of those skilled in the art, the application will be further described below in conjunction with the embodiments and the drawings, and the content mentioned in the embodiments is not a limitation of the application.

[0053] Referring to Figure 1 The static obstacle detection system based on ultrasonic waves of the application, comprising: a static obstacle detection device and a central server, the static obstacle detection device is installed in a place where static obstacle detection is needed, and is used for obstacle detection; the central server is used for receiving, storing and managing the data sent by the static obstacle detection device, and sends an alarm notification when there is a static obstacle;

[0054] Referring to Figure 2As shown, in the preferred example, the static obstacle detection device comprises: a microcomputer 1, a microphone array 2, a speaker array 3, a power amplifier module 4, a 4G wireless communication module 5, and a device shell 6.

[0055] The microcomputer is used to control the speaker array to emit ultrasonic waves and the microphone array to receive the ultrasonic wave signals reflected back by the obstacles, and process the received reflected signals to determine whether a static obstacle exists, and upload the result data to the central server through the network.

[0056] In the preferred example, the speaker array has one or more speakers connected to the power amplifier module through a line and then connected to the audio socket of the microcomputer, and emits ultrasonic waves with a specific waveform.

[0057] In the preferred example, the microphone array has one or more omnidirectional microphones with a recording frequency range of 20Hz-20kHz and a sampling rate of up to 48kHz, and the microphone array is connected to the microcomputer through USB, serial port or network port to receive the ultrasonic wave signals reflected back by the obstacles.

[0058] In the preferred example, the ultrasonic waves emitted by the speaker array are composed of linear frequency modulation waves (FMCW) with a frequency range of 15-22kHz.

[0059] In the preferred example, the central server provides a user interface through the network, has a device list and a system log list to display the status of all devices, provides device management functions, and can remind the system administrator to pay attention through a warning popup after the device detects an obstacle.

[0060] Reference Figure 3 As shown, the static obstacle detection method based on ultrasonic waves of the present application is based on the above system and comprises the following steps:

[0061] (1) Initialization: obtaining an initial detection result y0(d) through an environment detection algorithm in an open environment, where d is the distance between the obstacle and the device, and the initial detection result is a curve with distance as the horizontal coordinate and obstacle intensity as the vertical coordinate;

[0062] (2) Continuously running the environment detection algorithm at a certain time interval, and obtaining a real-time detection result y k (d) each time;

[0063] (3) Setting an adaptive detection threshold Y(d) = Y0 / d 2 , where Y0 is an empirical value and takes a value of (0, 1) (because the energy of ultrasonic waves decays inversely as the square of the distance, i.e., the farther the distance, the weaker the signal, so the threshold should be lower);

[0064] (4) Real-time detection result y k (d) Difference with initial detection result y0(d), if there is a distance d, such that y k (d) -y0(d)>Y(d), that is, if it exceeds the adaptive detection threshold after the difference, it is determined that there is a new static obstacle at the distance d;

[0065] (5) According to the static obstacle filtering algorithm, the detected static obstacles are filtered, and the static obstacles meeting the conditions are sent to the central server.

[0066] Among them, the open environment refers to the initial situation of the target place to be detected without obstacles, and all building structures and objects in this environment will be regarded as background and not detected as obstacles.

[0067] More specifically, referring to Figure 4 , the environment detection algorithm is specifically:

[0068] (a) Generate a signal to be sent: the signal to be sent is composed of N frames, each frame contains a linear frequency modulation wave wav sig with a time length of t sig and a blank time of t int , wherein the frequency of the linear frequency modulation wave wav sig is f min ~f max (the waveform and spectrum of the signal to be sent are shown in Figure 5 );

[0069] (b) Send and receive ultrasonic waves: play the signal to be sent through the loudspeaker array, and record the audio signal received by the microphone array as wav recv ;

[0070] (c) Filter noise: pass the received audio signal through a band-pass filter with a frequency range of f min ~f max to filter out noise interference of other frequency spectra;

[0071] (d) Calculate correlation: calculate the normalized cross-correlation function ccf of the above filtered audio signal and the linear frequency modulation wave wav sig , and the value range of the normalized cross-correlation function is [0, 1];

[0072] (e) Multi-cycle mean filtering: and perform peak detection on the normalized cross-correlation function ccf for a maximum of N points, and according to the peak points p1, p2,..., p n , divide and restore the normalized cross-correlation function ccf into N frames, and to reduce the influence of Gaussian noise dynamic targets, superimpose and average the N frames to perform multi-cycle mean filtering;

[0073] (f) extracting the upper envelope of the above mean filtered curve to eliminate the phase fluctuation of the waveform, and obtaining a stable environmental detection result y(d).

[0074] wherein the function expression of the linear frequency modulation wave is:

[0075]

[0076] wherein A is the amplitude of the signal, is the center frequency of the signal, B=F h -F l is the bandwidth of the sweep, F h and F l are the highest and lowest scanning frequencies, t is time, T is the scanning time, and φ0 is the initial phase.

[0077] wherein the calculation formula of the cross-correlation function is:

[0078]

[0079] wherein f1(t) and f2(t) are two function sequences with time as the independent variable, t is time, and τ is the integral variable; the normalization means scaling the data to 0 and 1 according to the maximum value and the minimum value, and the formula is: x max is the maximum value of the data, and x min is the minimum value of the data.

[0080] wherein the static obstacle filtering algorithm means that the obstacle needs to meet the following conditions to report to the center server, otherwise it is determined to be a pedestrian or environmental interference:

[0081] (51) if the static obstacle detection device contains multiple microphones, and a certain number of microphones detect the obstacle (the parameter can be adjusted, and generally can be set to 1 / 2);

[0082] (52) the obstacle exists for a certain time (the parameter can be adjusted, and generally can be set to one minute).

[0083] The present application has many specific application approaches, and the above description is only the preferred embodiment of the present application. It should be pointed out that for ordinary skilled persons in the technical field, some improvements can be made without departing from the principle of the present application, and these improvements should also be regarded as the protection scope of the present application.

Claims

1. An ultrasonic based static obstacle detection method based on an ultrasonic based static obstacle detection system, the system comprising: Static obstacle detection device and central server, the static obstacle detection device is installed in the place where static obstacle detection is needed, for obstacle detection; The central server receives, stores and manages the data sent by the static obstacle detection device, and sends an alarm notification when there is a static obstacle; The static obstacle detection device comprises a microcomputer, a microphone array and a speaker array; The microcomputer controls the speaker array to emit ultrasonic waves and the microphone array to receive the reflected ultrasonic wave signals, and processes the received reflected signals to determine whether a static obstacle exists, and uploads the result data to the central server through the network; The method comprises the following steps: (1) Initialization: obtaining an initial detection result y0(d) in an open environment by an environment detection algorithm, d is the distance between the obstacle and the device, and the initial detection result is a curve with distance as the horizontal coordinate and obstacle intensity as the vertical coordinate; (2) Real-time detection: According to a certain time interval, the environment detection algorithm is continuously run, and each time a real-time detection result y is obtained k (d); (3) Threshold setting: set adaptive detection threshold Y(d) = Y0 / d 2 Y0 is an empirical value, taking value (0, 1); (4) Difference analysis: the real-time detection result y k (d) Difference with the initial detection result y0(d), if there is a distance d, such that y k (d) -y0(d) > Y(d), that is, if it exceeds the adaptive detection threshold after difference, it is determined that there is a newly added static obstacle at the distance d; (5) Filtering the detected static obstacles according to a static obstacle filtering algorithm, and sending the static obstacles meeting the conditions to the central server; The open environment refers to the initial situation of the target place to be detected without obstacles, in which all building structures and objects are regarded as background and are not detected as obstacles; The environment detection algorithm is as follows: (a) generating a signal to be transmitted: the signal to be transmitted is composed of N frames, each frame containing a linear frequency modulation wave wav sig with a duration of t sig and a blanking of t int seconds, wherein the linear frequency modulation wave wav sig has a frequency of f min ~ f max ; (b) Transmit and receive ultrasound waves: Play the signal to be transmitted through the speaker array, and record the audio signal received by the microphone array as wav recv ; (c) filtering out noise: passing the received audio signal through a band-pass filter with a frequency range of f min ~ f max to filter out noise interference from other frequency spectrums; (d) calculating the correlation: calculating the normalized cross-correlation function ccf of the above filtered audio signal and the chirp wav sig , the normalized cross-correlation function has a value range of [0, 1]; (e) Multi-cycle mean filtering: peak detection of the normalized cross-correlation function ccf with a maximum of N points, according to the peak points p1, p2,..., pN, the mean value of the peak points is taken as the final result. n The normalized cross-correlation function ccf is divided into N frames, and the N frames are superimposed to take the mean value, and the multi-cycle is subjected to mean filtering. (f) Extracting the upper envelope of the above-mentioned mean filtered curve to eliminate the phase fluctuation of the waveform and obtain a stable environment detection result y(d); The static obstacle filtering algorithm filters the detected static obstacles, i.e. the obstacle needs to meet the following conditions to report to the central server, otherwise it is judged as a pedestrian or environmental interference: (51) If the static obstacle detection device contains multiple microphones, a certain number of microphones detect the obstacle; (52) The obstacle exists for a certain period of time.

2. The ultrasonic-based static obstacle detection method according to claim 1, characterized in that, The function expression of the linear frequency modulation wave is: where A is the amplitude of the signal, F is the center frequency of the signal, B = F h F l F is the bandwidth of the sweep, h F l F are the highest and lowest sweep frequencies, t is time, T is the sweep time, and φ0 is the initial phase.

3. The ultrasonic-based static obstacle detection method of claim 1, wherein, The calculation formula of the cross-correlation function is: where f1(t) and f2(t) are two time-dependent function sequences, t is time, and τ is the integration variable; normalization means scaling the data between 0 and 1 according to the maximum and minimum values, and the formula is: x max is the maximum value of the data, and x min is the minimum value of the data.

Images