Method for detecting metal fragments falling off on a sauce filling production line

Abstract

The present application relates to the technical field of sauce filling production line, and discloses a detection method for metal fragments falling off on a sauce filling production line, comprising the following steps: S1, arranging solid-state laser detection radar, arranging multiple groups of solid-state laser detection radar on the conveying device of the filling production line, and performing laser detection on the sauce bottles moving quickly on the filling production line. The detection method for metal fragments falling off on the sauce filling production line uses the FLASH technology of the solid-state laser detection radar and the millimeter wave radar to monitor the change of the shape of the friction and vulnerable metal equipment, to determine whether metal fragments fall off, and uses a fusion algorithm to detect whether the detection results of the two radars match, including whether the size of the friction falling metal shape matches and whether the position of the metal fragments falling matches, so that the fault of the filling equipment can be found in time, the sauce bottles mixed with metal fragments can be removed, and harm to the human body can be avoided, which has practical significance and economic benefits.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sauce filling production line, in particular to a detection method for metal fragments falling off on a sauce filling production line. BACKGROUND

[0002] Sauce is a paste condiment processed from beans, wheat flour, fruits, meat or fish, etc. It originated in China and has a long history. Commonly seen sauce in China is divided into two categories: sweet sauce made of wheat flour and bean sauce made of beans; and other condiments such as meat sauce, fish sauce and fruit sauce. With the progress of sauce making technology, the method of making sauce is also used for cooking other non-sauce dishes, gradually developing a method of cooking dishes. Sauce needs to be filled into sauce bottles during production and processing.

[0003] However, the filled sauce bottles are not transparent, and the sauce is also not transparent. Due to the high speed of the filling equipment, the high production efficiency, the mechanical friction part and the gear transmission part of the filling production line, metal fragments falling off due to friction are mixed into the sauce bottles. The existing detection equipment and method are inefficient, cannot quickly detect, are not suitable for high-speed production lines, and have large detection errors. SUMMARY

[0004] (I) Technical problems solved

[0005] In view of the deficiencies in the prior art, the present application provides a detection method for metal fragments falling off on a sauce filling production line.

[0006] (II) Technical solutions

[0007] To achieve the above-mentioned purpose, the present application provides the following technical solutions: a detection method for metal fragments falling off on a sauce filling production line, comprising the following steps:

[0008] S1, arranging solid-state laser detection radar

[0009] On the conveying device of the filling production line, multiple groups of solid-state laser radars are arranged to perform laser detection on the sauce bottles moving quickly on the filling production line. The FLASH technology of the solid-state laser radar is used to detect the abnormal point cloud part with strong reflection in the laser point cloud, and measure the size and position of the abnormal point cloud shape.

[0010] S2, constructing a constant false alarm algorithm model

[0011] The constant false alarm rate algorithm based on random sampling is used for solid-state laser radar moving target detection. The constant false alarm rate algorithm based on random sampling is to simulate Monte Carlo independent random experiments to obtain unknown characteristic value estimation by randomly sampling a two-dimensional range-doppler matrix (RDM) in the data processing process of the solid-state laser radar according to the Monte Carlo principle. The estimation of the target background noise at the current time is obtained by processing the sample points of random sampling. The target decision threshold is obtained by using the noise estimation value, and the current moving target detection is finally realized.

[0012] S3, arranging a millimeter wave radar

[0013] On the metal filling equipment, a plurality of millimeter wave radars are arranged. For the high-speed running frictional metal filling equipment part, the millimeter wave radar is used to monitor the change of the shape of the friction and the fragile metal equipment to judge whether the metal debris falls.

[0014] S4, constructing a millimeter wave radar algorithm model

[0015] The direction finding area is divided into N parts in the horizontal direction and M parts in the vertical direction to obtain MxN grids. The intensity change of the signal emitted by the radar transmitting unit to each grid of the direction finding area is analyzed in turn to obtain the propagation function H1, and the intensity change of the echo signal generated by each grid to the R radar receiving units is analyzed to obtain the echo function H2. A control matrix A containing Q control functions corresponding to the Q radar transmitting units is set, and the signals emitted by the Q radar transmitting units are controlled respectively according to the control matrix A, so that the Q radar transmitting units emit the emitted signals projected to the direction finding area. The R radar receiving units are used to receive the echo signals returned from the direction finding area, and the echo signals received by the R radar receiving units and the echo matrix y formed by the echo signals are obtained. The reflection coefficients of each grid of the direction finding area are calculated according to the propagation function H1, the echo function H2, the control matrix A and the echo matrix y.

[0016] S5, calculating the millimeter wave radar detection result

[0017] The reflection coefficients of the calculated grids are back calculated to obtain the reflection coefficient matrix x represented by the MxN matrix. Whether each reflection coefficient in the reflection coefficient matrix x is greater than a predetermined threshold is judged in turn. When it is greater, it is determined that there is an obstacle in the corresponding direction of the corresponding grid.

[0018] S6, detection result identification

[0019] The detection data of the two kinds of radar detection are compared by using the new fusion algorithm to detect whether the detection results of the two kinds of radars match, including whether the size of the friction and the shape of the metal falling agree, and whether the position of the metal debris falling matches.

[0020] Preferably, the multiple groups of solid-state laser radars in S1 are respectively installed on both sides of the conveying device, and the detection directions of the solid-state laser radars on both sides of the conveying device are concentrated on the positions of the sauce bottles on the conveying belt.

[0021] Preferably, the constant false alarm rate algorithm model based on random sampling is constructed in S2, parameters of the constant false alarm rate algorithm based on random sampling are determined, random sampling is performed on the entire radar detection area based on the determined parameters, background noise estimation and decision threshold of radar detection are determined, and radar multi-target detection is performed.

[0022] Preferably, the parameters of the constant false alarm rate algorithm include sampling points and threshold factors, and the parameters of the constant false alarm rate algorithm based on random sampling are determined by adopting a Monte Carlo method to determine the values of the sampling points and the threshold factors.

[0023] Preferably, the method for obtaining the propagation function H1 in S4 is an analytical derivation method, including the following steps:

[0024] 1) An analog object surface parallel to the plane where the radar transmitting units are located is set between the direction-finding area and the transmitting radar, and the analog object surface is divided into MxN grids.

[0025] 2) The radar transmitting units transmit test signals to the analog object surface, and the analog signal receiving radar receives the test signals on the analog object surface, so as to obtain actual signals after the test signals are transmitted and propagated to the analog object surface.

[0026] Preferably, the strength relationship between the test signals and the actual signals is simulated and derived to obtain the propagation function H1.

[0027] Preferably, the Q control functions in the control matrix A in S4 respectively perform amplitude modulation, frequency modulation and phase modulation on the transmission signals of the Q radar transmitting units, so that the Q radar transmitting units transmit the transmission signals.

[0028] Preferably, the constructed constant false alarm rate algorithm model based on random sampling determines the background noise estimation and the threshold factor of radar detection and determines the decision threshold.

[0029] (Three) beneficial effects

[0030] Compared with the prior art, the present application provides a detection method for metal fragments falling off on a sauce filling production line, which has the following beneficial effects:

[0031] The sauce filling production line metal debris detection method, through the filling production line fast moving sauce bottle, using solid state laser radar FLASH technology, in the laser point cloud detects the reflection strong abnormal point cloud part, and measures the abnormal point cloud shape size and position, and for the high speed running friction metal filling equipment part, using millimeter wave radar monitoring friction and easy to wear metal equipment shape change, to judge whether the metal debris falls, through the fusion algorithm, to detect whether the two radar detection results match, including whether the friction metal shape size matches, whether the metal debris falling position matches, can timely find the filling equipment failure, eliminate the mixed metal debris sauce bottle, avoid the harm to the human body, has practical significance and economic benefit. BRIEF DESCRIPTION OF DRAWINGS

[0032] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, illustrate the present application, and are used to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0033] Figure 1 The flowchart of the present application. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments.

[0035] Examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation on the present application.

[0036] As Figure 1 shown, the present application provides a sauce filling production line metal debris detection method, comprising the following steps:

[0037] S1, arranging solid state laser detection radar

[0038] On the conveying device of the filling production line, a plurality of solid state laser radars are arranged, and laser detection is performed on the fast moving sauce bottle on the filling production line. FLASH technology of the solid state laser radar is used to detect the reflection strong abnormal point cloud part in the laser point cloud, and measure the abnormal point cloud shape size and position. The plurality of solid state laser radars are respectively installed on both sides of the conveying device, and the detection directions of the solid state laser radars on both sides of the conveying device are gathered at the position of the sauce bottle on the conveying belt.

[0039] S2, constructing a constant false alarm algorithm model

[0040] The constant false alarm rate algorithm based on random sampling is used for solid-state laser radar moving target detection. The constant false alarm rate algorithm based on random sampling is to simulate Monte Carlo independent random experiments to obtain unknown characteristic value estimation by randomly sampling a two-dimensional range-doppler matrix (RDM) in the data processing process of the solid-state laser radar according to the Monte Carlo principle. The estimation of the target background noise at the current time is obtained by processing the sample points of random sampling. The target decision threshold is obtained by using the noise estimation value, and the current moving target detection is finally realized. A constant false alarm rate algorithm model based on random sampling is constructed, and the parameters of the constant false alarm rate algorithm based on random sampling are determined. The background noise estimation value and the decision threshold of the radar detection are determined by randomly sampling the entire radar detection area based on the determined parameters, and the radar multi-target detection is performed.

[0041] The parameters of the constant false alarm rate algorithm include sample points and threshold factors. The parameters of the constant false alarm rate algorithm based on random sampling include the values of the sample points and the threshold factors determined by the Monte Carlo method. The constant false alarm rate algorithm model based on random sampling determines the background noise estimation value and the threshold factor of the radar detection and determines the decision threshold.

[0042] S3, arranging a millimeter wave radar

[0043] On the metal filling equipment, a plurality of millimeter wave radars are arranged. For the high-speed running frictional metal filling equipment part, the millimeter wave radar is used to monitor the change of the shape of the friction and the fragile metal equipment to judge whether the metal fragments fall off.

[0044] S4, constructing a millimeter wave radar algorithm model

[0045] The direction finding area is divided into N parts in the horizontal direction and M parts in the vertical direction to obtain MxN grids. The intensity change of the signal emitted by the radar transmitting unit to each grid of the direction finding area is analyzed in turn to obtain the propagation function H1, and the intensity change of the echo signal generated by each grid to the R radar receiving units is analyzed to obtain the echo function H2. A control matrix A containing Q control functions corresponding to the Q radar transmitting units is set, and the signals emitted by the Q radar transmitting units are controlled respectively according to the control matrix A, so that the Q radar transmitting units emit the emitted signals projected to the direction finding area. The R radar receiving units are used to receive the echo signals returned from the direction finding area, and the echo matrix y formed by the echo signals received by the R radar receiving units is obtained. The reflection coefficients of each grid of the direction finding area are calculated according to the propagation function H1, the echo function H2, the control matrix A, and the echo matrix y.

[0046] The method for obtaining the propagation function H1 is an analysis and deduction method, which includes the following steps:

[0047] 1) Set an analog object plane parallel to the plane where the transmitting radar unit is located between the direction finding area and the transmitting radar, and divide the analog object plane into MxN grids.

[0048] 2) Transmit a test signal to the analog object plane using the radar transmitting unit, and receive the actual signal after the test signal is transmitted and propagated to the analog object plane using the analog signal receiving radar, thereby obtaining the actual signal.

[0049] According to the strength relationship between the test signal and the actual signal, the propagation function H1 is obtained by analog deduction, and the Q control functions in the control matrix A are used to adjust the amplitude, frequency and phase of the transmission signals of the Q radar transmitting units, so that the Q radar transmitting units transmit the transmission signals.

[0050] S5, calculate the millimeter wave radar detection result

[0051] The reflection coefficient matrix x represented by the MxN matrix is obtained by backstepping the calculated reflection coefficient of the grid, and it is judged in turn whether each reflection coefficient in the reflection coefficient matrix x is greater than a predetermined threshold value, and when it is greater than the predetermined threshold value, it is determined that there is an obstacle in the corresponding direction of the corresponding grid.

[0052] S6, detection result identification

[0053] Using a new fusion algorithm, the detection data of the two kinds of radar detection are compared to detect whether the detection results of the two kinds of radar detection match, including whether the size of the friction falling metal shape matches, and whether the position of the metal debris falling matches.

[0054] The working principle of the detection method of the metal debris falling on the sauce filling production line will be described in detail below.

[0055] By aiming at the fast-moving sauce bottles on the filling production line, using the FLASH technology of solid-state laser radar, the abnormal point cloud part with strong reflection in the laser point cloud is detected, and the size and position of the abnormal point cloud shape are measured, and for the high-speed running and friction metal filling equipment part, the millimeter wave radar is used to monitor the change of the shape of the friction and vulnerable metal equipment to judge whether the metal debris falls, by using the fusion algorithm, whether the detection results of the two kinds of radar detection match is detected, including whether the size of the friction falling metal shape matches, and whether the position of the metal debris falling matches, the fault of the filling equipment can be found in time, the sauce bottles mixed with metal debris are rejected, and the harm to the human body is avoided, which has practical significance and economic benefit.

Claims

1. A method of detecting dislodged metal fragments on a sauce filling production line, characterized in that, Comprise the following steps: S1, arrange solid-state laser detection radar On the conveying device of the filling production line, a plurality of solid-state laser radars are arranged, which are used for laser detection of the fast-moving sauce bottle on the filling production line, and the FLASH technology of the solid-state laser radar is used to detect the abnormal point cloud part with strong reflection in the laser point cloud, and the size and position of the abnormal point cloud shape are measured; S2, construct constant false alarm algorithm model The constant false alarm algorithm based on random sampling is used for solid-state laser radar moving target detection. The constant false alarm algorithm based on random sampling is to simulate Monte Carlo independent random experiment to obtain unknown characteristic estimation process by randomly sampling two-dimensional range-doppler matrix (RDM) in solid-state laser radar data processing process according to Monte Carlo principle. The estimation of target background noise at the current time is obtained by processing the sample points of random sampling. The target decision threshold is obtained by using noise estimation value to finally realize current moving target detection; S3, arrange millimeter wave radar On the metal filling equipment, a plurality of millimeter wave radars are arranged, which are used for monitoring the change of the shape of the friction and vulnerable metal equipment to judge whether the metal debris falls off; S4, construct millimeter wave radar algorithm model The detection area is divided into N parts in the horizontal direction and M parts in the vertical direction to obtain M×N grids. The intensity change of the signal emitted by the radar transmitting unit to each grid in the detection area is analyzed in turn to obtain the propagation function H1, and the intensity change of the echo signal generated by each grid to the R radar receiving units is analyzed to obtain the echo function H2. The control matrix A containing Q control functions corresponding to the Q radar transmitting units is set, and the signals emitted by the Q radar transmitting units are controlled respectively according to the control matrix A, so that the Q radar transmitting units emit the emitted signals projected to the detection area. The R radar receiving units receive the echo signals returned from the detection area to obtain the echo matrix y formed by the echo signals received by the R radar receiving units respectively. The reflection coefficients of each grid in the detection area are calculated according to the propagation function H1, the echo function H2, the control matrix A and the echo matrix y; S5, calculate the millimeter wave radar detection result The reflection coefficients of the calculated grids are back calculated to obtain the reflection coefficient matrix x represented by M×N matrix. Whether each reflection coefficient in the reflection coefficient matrix x is greater than a predetermined threshold is judged in turn. When it is greater, it is determined that there is an obstacle in the corresponding direction of the corresponding grid; S6, detection result identification The detection data of the two kinds of radar detection are compared by using the new fusion algorithm to detect whether the detection results of the two kinds of radar match, including whether the friction and the size of the metal shape match, and whether the position of the metal debris falling matches.

2. The method of claim 1, wherein the method comprises: The plurality of solid-state laser radars in S1 are respectively installed on both sides of the conveying device, and the detection directions of the solid-state laser radars on both sides of the conveying device are gathered at the position of the sauce bottle on the conveying belt.

3. The method of claim 1, wherein the method comprises: The S2 constructs a constant false alarm rate algorithm model based on random sampling, and determines parameters of the constant false alarm rate algorithm based on random sampling, uses the determined parameters to randomly sample the whole radar detection area, determines a background noise estimation value and a decision threshold of radar detection, and performs radar multi-target detection.

4. The method of claim 3, wherein the method further comprises: determining the presence of the metal fragment on the sauce filling production line based on the detected signal. The parameters of the constant false alarm rate algorithm include sampling points and a threshold factor, and the determination of the parameters of the constant false alarm rate algorithm based on random sampling includes determining the values of the sampling points and the threshold factor by using a Monte Carlo method. ​ 5. The method of claim 1, wherein the method comprises: The method for obtaining the propagation function H1 in the S4 is an analytical derivation method, including the following steps: ​ 1) setting an analog object surface parallel to the plane where the radar units are located between the direction finding area and the transmitting radar, and dividing the analog object surface into M×N grids; 2) transmitting a test signal from the radar transmitting units to the analog object surface, and receiving the test signal at the analog object surface by using the analog signal receiving radar, so as to obtain the actual signal after the test signal is transmitted and propagated to the analog object surface.

6. The method for detecting metal fragments falling off a sauce filling production line according to claim 5, characterized in that: The propagation function H1 is obtained by analog derivation according to the intensity relationship between the test signal and the actual signal.

7. The method of claim 1, wherein the method comprises: The Q control functions in the control matrix A respectively perform amplitude modulation, frequency modulation and phase modulation on the transmitting signals of the Q radar transmitting units, so that the Q radar transmitting units transmit the transmitting signals.

8. The method of claim 3, wherein the method is characterized by: The constructed constant false alarm rate algorithm model based on random sampling determines the background noise estimation value and the threshold factor of radar detection, and determines the decision threshold.

Images