Radar signal real-time processing method and system based on FPGA

Abstract

The invention provides a radar signal real-time processing method and system based on an FPGA, and the method comprises the steps: carrying out the region determination based on the moving trend of a corresponding moving person located in a target space, and obtaining a person moving region; in response to a region overlapping relationship between the equipment moving region of any space equipment and the personnel moving region, determining the space equipment as having an interference attribute, and extracting equipment point cloud data corresponding to the space equipment having the interference attribute and personnel point cloud data corresponding to the moving personnel; performing track comparison on a personnel moving track obtained by track prediction based on the personnel point cloud data and an equipment moving track obtained by track prediction based on the equipment point cloud data; and determining that an interference phenomenon exists between the space equipment and the mobile personnel at any track moment, and sending early warning information formed by the track moment and interference data generated corresponding to the interference phenomenon to the mobile terminal. The early warning accuracy is at least improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a radar signal real-time processing method and system based on FPGA. BACKGROUND

[0002] With the rapid development of industrial automation and intelligence, the interaction between personnel and equipment in industrial scenarios is becoming more and more frequent. How to protect personnel safety and avoid conflicts between equipment operation and personnel activities has become an important issue in the field of industrial safety production.

[0003] Existing radar-based personnel and equipment detection technologies mostly use simple region division and collision detection algorithms to determine whether personnel and equipment are in the same region, and then issue an alarm. However, these technologies mainly focus on whether personnel and equipment enter a pre-set dangerous region, and lack in-depth analysis of the dynamic changes of equipment operation trajectories and personnel movement trajectories, making it impossible to achieve accurate early warning of potential collision risks.

[0004] In actual industrial scenarios, the operation trajectories of industrial equipment are complex and variable, and employees also frequently move when performing tasks. Existing technologies cannot effectively warn and remind employees according to the real-time operation trajectories of industrial equipment, which leads to difficulties for employees to timely predict the movement trend of equipment during equipment operation, resulting in a high security risk and failing to meet the high requirements of personnel and equipment safety interaction in industrial scenarios.

[0005] Therefore, there is an urgent need for a radar signal real-time processing method and system based on FPGA that can improve the accuracy of early warning. SUMMARY

[0006] Based on the above problems, the present application is proposed to provide a radar signal real-time processing method and system of FPGA that overcomes the above problems or at least partially solves the above problems.

[0007] According to one aspect of the present application, a radar signal real-time processing method based on FPGA is provided, comprising the following steps:

[0008] Based on the movement trend of the corresponding moving personnel located in the target space, a region determination is performed to obtain a personnel movement region corresponding to the moving personnel;

[0009] In response to the existence of a region overlap relationship between the device movement region of any spatial device located in the target space and the personnel movement region, the spatial device is determined to have an interference attribute, and based on the obtained radar signal, device point cloud data of the spatial device with the interference attribute closest to the moving personnel and personnel point cloud data of the moving personnel are extracted;

[0010] The trajectory of the moving person is predicted based on the personnel point cloud data of the moving person, and the trajectory of the device is predicted based on the device point cloud data of the space device.

[0011] In response to determining that the space device and the moving person interfere with each other at any trajectory moment based on the comparison result, pre-warning information composed of interference data generated from the trajectory moment and the interference phenomenon corresponding to the interference phenomenon is sent to a mobile terminal corresponding to the moving person.

[0012] Optionally, in the method according to the present application, the personnel moving area corresponding to the moving person in the target space is determined based on the moving trend of the moving person in the target space, and the personnel moving area corresponding to the moving person is obtained, comprising:

[0013] The space route and the device sub-image indicating each space device in the target space are determined based on the space view corresponding to the target space.

[0014] The terminal position and the moving direction of the mobile terminal corresponding to the moving person are determined based on the moving positioning signal sent by the mobile terminal corresponding to the moving person in the space view.

[0015] The space route is divided based on the terminal position, and a first divided route in an inverse relationship with the moving direction and a second divided route in a positive relationship are obtained.

[0016] In response to the second divided route not having any branch intersection, the personnel moving area corresponding to the moving person is determined based on the second divided route.

[0017] In response to the second divided route having any branch intersection, each space device indicated by each device sub-image is determined as a first device when the first divided route passes through the device sub-image, and each space device indicated by each device sub-image is determined as a second device when the second divided route passes through the device sub-image.

[0018] Each second device is predicted to be interacted with based on the device interaction of the moving person corresponding to the first divided route, and the personnel moving area corresponding to the moving person is determined based on the second device having an interaction probability.

[0019] Optionally, in the method according to the present application, each second device is predicted to be interacted with based on the current interaction record of the moving person corresponding to the first divided route, and the personnel moving area corresponding to the moving person is determined based on the second device having an interaction probability, comprising:

[0020] Each first device interacted with by the moving person is input into an interaction prediction model trained in advance to obtain an interaction probability corresponding to each second device.

[0021] In response to any second device existing interaction probability corresponding to a plurality of first devices, all interaction probabilities corresponding to the second device are fused to obtain an updated interaction probability;

[0022] The second device corresponding to the existing interaction probability is determined as a preliminary interaction device, and a preliminary sub-route from the initial personnel position to each preliminary interaction device is determined based on the second divided route;

[0023] A route selection interface is created, and each preliminary sub-route arranged in descending order of corresponding interaction probability is filled into each route slot arranged in ascending order in the route selection interface;

[0024] In response to the mobile terminal selecting any route slot, a personnel movement area corresponding to the mobile personnel is determined based on the preliminary sub-route filled in the route slot.

[0025] Optionally, in the method according to the present application, all interaction probabilities corresponding to the second device are fused to obtain an updated interaction probability, comprising:

[0026] The number of probabilities corresponding to all interaction probabilities is obtained, and a number evaluation value corresponding to the number dimension is determined based on the number of probabilities;

[0027] All interaction probabilities are mean calculated, and a mean evaluation value corresponding to the mean dimension is determined based on the obtained probability;

[0028] The number evaluation value and the mean evaluation value are weighted and summed, and a probability growth rate is determined based on the obtained comprehensive evaluation value;

[0029] The corresponding maximum of all interaction probabilities is updated by the probability growth rate to obtain an updated interaction probability.

[0030] Optionally, in the method according to the present application, in response to any spatial device in the target space having a region overlap relationship with the personnel movement area, the spatial device is determined to have an interference attribute, and based on the obtained radar signal, the device point cloud data of the spatial device corresponding to the mobile personnel and the personnel point cloud data corresponding to the mobile personnel are extracted, comprising:

[0031] A target twin space corresponding to the target space is created, and based on the coordinate processing of the target twin space, each device edge point at the edge position of each device movement area and each personnel edge point at the edge position of the personnel movement area are determined;

[0032] The edge point of each device and the edge point of each person are multiplied by the preset inflation coefficient, and each device inflation model and person inflation model in the target twin space are determined based on the obtained device inflation point and person inflation point;

[0033] In response to any device inflation model and the person inflation model having a region overlap relationship, the spatial device corresponding to the device inflation model is determined to have an interference attribute;

[0034] The first distance corresponding to each device inflation model and the person inflation model is determined, and the device point cloud data of the spatial device and the person point cloud data of the moving person corresponding to the minimum first distance are extracted based on the radar detection unit.

[0035] Optionally, in the method according to the present application, the trajectory comparison between the person moving trajectory obtained based on the trajectory prediction of the corresponding person moving area based on the person point cloud data and the device moving trajectory obtained based on the trajectory prediction of the corresponding device moving area based on the device point cloud data comprises:

[0036] The initial person position, the initial person shape and the person speed of the corresponding moving person are determined based on the person point cloud data, and the initial device position, the initial device shape and the device speed of the corresponding spatial device are determined based on the device point cloud data;

[0037] The moving length of the corresponding person moving area is obtained, and the trajectory time period corresponding to the moving length is determined based on the person speed, wherein the trajectory time period comprises different trajectory time points;

[0038] The current person position of the moving person corresponding to each trajectory time point is determined based on the initial person position, and the initial person shape is changed in shape corresponding to each trajectory time point based on the person shape change rule corresponding to the moving person, to obtain the current person shape;

[0039] The current device position of the spatial device corresponding to each trajectory time point is determined based on the initial device position, and the initial device shape is changed in shape corresponding to each trajectory time point based on the device shape change rule corresponding to the spatial device, to obtain the current device shape;

[0040] The person moving trajectory composed of the current person shape and the current person position corresponding to the same trajectory time point is compared with the device moving trajectory composed of the current device shape and the current device position corresponding to the same trajectory time point.

[0041] Optionally, in the method according to the present application, the trajectory comparison between the person moving trajectory composed of the current person shape and the current person position corresponding to the same trajectory time point and the device moving trajectory composed of the current device shape and the current device position corresponding to the same trajectory time point comprises:

[0042] determining a first position point and a second position point in the target twin space based on the current personnel position and the current device position corresponding to the same trajectory time;

[0043] generating a current personnel model corresponding to the current personnel posture with the first position point as the center and generating a current device model corresponding to the current device posture with the second position point as the center;

[0044] in response to the current device model and the current personnel model having a model overlap relationship, determining that the spatial device corresponding to the current device model and the moving personnel have an interference phenomenon;

[0045] in response to the current device model and the current personnel model not having a model overlap relationship, obtaining a second distance between the current device model and the current personnel model, and calling a device attribute label of the spatial device corresponding to the current device model;

[0046] determining a safety distance corresponding to the spatial device based on the device attribute label, and in response to the second distance being less than or equal to the safety distance, determining that the spatial device corresponding to the current device model and the moving personnel have an interference phenomenon, and otherwise determining that there is no interference phenomenon.

[0047] Optionally, in the method according to the present application, in response to determining that the spatial device and the moving personnel have an interference phenomenon at any trajectory time based on the comparison result, pre-warning information composed of the trajectory time and the interference phenomenon corresponding to the interference phenomenon is sent to a mobile terminal corresponding to the moving personnel, including:

[0048] in response to determining that any spatial device and the moving personnel have an interference phenomenon at any trajectory time based on the comparison result, determining the trajectory time as a starting time, and collecting all other trajectory times after the starting time to an avoidance selection group;

[0049] generating each future device model corresponding to the current device posture at each trajectory time in the avoidance selection group with the second position point as the center, and determining a trajectory time corresponding to no interference phenomenon with the moving personnel and closest to the starting time as an avoidance time;

[0050] calling a preset pre-warning template, and filling device information corresponding to the spatial device and an avoidance period composed of the trajectory time and the avoidance time into a device slot and a period slot in the preset pre-warning template to obtain pre-warning information sent to a mobile terminal corresponding to the moving personnel.

[0051] Optionally, in the method according to the present application, the method further comprises:

[0052] In response to the early warning information sent to the mobile terminal, an ambient light sensing unit and a proximity sensing unit pre-set in the mobile terminal are determined, and a timing task lasting for a preset time period is established;

[0053] In response to the light collection value output by the ambient light sensing unit being less than a preset light threshold value and the proximity sensing unit not triggering a human proximity signal based on the timing task, an early warning time point located before the starting time point and separated by a preset buffer time period is determined.

[0054] Personnel information corresponding to the moving personnel is acquired, and a safety number corresponding to a safety helmet worn by the moving personnel is determined based on the personnel information.

[0055] In response to reaching the early warning time point, a stimulation unit pre-set in the safety helmet corresponding to the safety number is controlled to output an electric stimulation signal.

[0056] According to another aspect of the present application, a radar signal real-time processing system based on FPGA is provided, comprising:

[0057] The region determination module is configured to determine a region based on the movement trend of the corresponding moving personnel located in the target space, to obtain a personnel movement region corresponding to the moving personnel.

[0058] The interference point cloud module is configured to determine a spatial device as having an interference attribute in response to the device movement region of any spatial device located in the target space and the personnel movement region having a region overlap relationship, and to extract device point cloud data of the spatial device having the interference attribute and closest to the moving personnel and personnel point cloud data corresponding to the moving personnel based on the acquired radar signal.

[0059] The trajectory comparison module is configured to compare a personnel movement trajectory obtained by predicting the trajectory of the corresponding personnel movement region based on the personnel point cloud data and a device movement trajectory obtained by predicting the trajectory of the corresponding device movement region based on the device point cloud data.

[0060] The interference early warning module is configured to send early warning information composed of interference data generated by the trajectory time point and the corresponding interference phenomenon to the mobile terminal corresponding to the moving personnel in response to determining that the spatial device and the moving personnel have an interference phenomenon at any trajectory time point based on the comparison result.

[0061] According to the scheme of the present application, first, the interference attribute of the equipment is judged in combination with the overlapping relationship between the equipment moving area and the personnel moving area by determining the personnel moving area based on the moving trend of the moving personnel, which can accurately identify the industrial equipment that may cause interference to the employees, avoiding the misjudgment and missed judgment problems caused by simple area division in the traditional technology; second, in the data processing link, the point cloud data of the equipment with the interference attribute closest to the moving personnel and the personnel point cloud data are extracted, which provides accurate original data support for subsequent trajectory prediction. Based on these point cloud data, the trajectories of the personnel and the equipment are predicted respectively, and the trajectories of the two are compared, which can analyze the change trend of the running trajectories of the personnel and the equipment in real time and dynamically. Compared with the defects of the prior art that lack in-depth analysis of the dynamic change of the trajectory, the present method can predict the potential collision risk in advance, realize accurate early warning of dangerous situations; finally, when it is detected that the personnel and the equipment have interference phenomenon at a certain trajectory moment, the early warning information containing the trajectory moment and the interference data is sent to the employee mobile terminal in time, and the employee can obtain the equipment running dynamics in the first time and make avoidance action in advance, which not only improves the prediction ability of the employee to the moving trend of the equipment, but also significantly reduces the safety hidden danger of the collision between the personnel and the equipment in the industrial scene, meets the high requirements of the personnel and the equipment for safe interaction in the industrial scene, and improves the corresponding early warning accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 A flow chart of a radar signal real-time processing method based on FPGA according to an embodiment of the present application is shown;

[0063] Figure 2 A structural block diagram of a radar signal real-time processing system based on FPGA according to another embodiment of the present application is shown. DETAILED DESCRIPTION

[0064] Exemplary embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art.

[0065] To solve the problems existing in the prior art, the inventors propose the scheme of the present application. An embodiment of the present application provides a radar signal real-time processing method based on FPGA, which can be executed in a computing device. The computing device can be understood as a terminal with data processing function, such as a mobile phone or a computer.

[0066] Figure 1 A flow chart of a radar signal real-time processing method based on FPGA according to an embodiment of the present application is shown, asFigure 1 As shown, the method starts from step S1, in which the following is included:

[0067] Based on the movement trend of the corresponding mobile personnel located in the target space, a personnel movement area corresponding to the mobile personnel is determined.

[0068] For example, the present embodiment can be applied in an industrial scene, aiming to remind employees to avoid according to the running track of industrial equipment. For the content of step S1, the present embodiment can first determine a region based on the movement trend of the mobile personnel in the target space to obtain the personnel movement area of the mobile personnel. It can be explained that in an industrial scene, the target space usually refers to an industrial work site such as a factory workshop or a storage area. By analyzing the movement trend of the mobile personnel in the space, the possible activity range can be determined more accurately. For example, the movement trend of the personnel can be determined in combination with historical movement data of the personnel, so as to determine the activity area of the personnel in the industrial space in advance, laying a foundation for subsequent comparison and analysis with the movement area of the industrial equipment, thereby more effectively predicting possible interference, improving the safety and coordination of personnel and equipment operation in the industrial scene, so that the activity range of each mobile personnel can be dynamically determined according to the specific situation of the mobile personnel, rather than using a fixed and unified region division method, thereby better adapting to the diversity and complexity of personnel movement in the industrial scene, further improving the accuracy and flexibility of personnel activity area determination, and providing a more reliable basis for subsequent interference detection and warning.

[0069] Further, in the present embodiment, the above-mentioned "determining a personnel movement area corresponding to the mobile personnel based on the movement trend of the corresponding mobile personnel located in the target space" can further include the following steps:

[0070] determining a space route and a device sub-image indicating each space device located in the target space based on the space-on-view corresponding to the target space;

[0071] determining a terminal position and a moving direction of the mobile terminal based on the movement positioning signal sent by the mobile terminal of the corresponding mobile personnel in the space-on-view;

[0072] dividing the space route based on the terminal position to obtain a first divided route in reverse relationship with the moving direction and a second divided route in positive relationship;

[0073] in response to the second divided route not existing any bifurcation intersection, determining a personnel movement area corresponding to the mobile personnel based on the second divided route;

[0074] in response to the second divided route existing any bifurcation, determining the spatial equipment indicated by each equipment sub-image of the first divided route as a first equipment, and determining the spatial equipment indicated by each equipment sub-image of the second divided route as a second equipment;

[0075] based on the equipment interaction of the moving person corresponding to the first divided route, making an interaction prediction on each second equipment, and determining a person moving area corresponding to the moving person based on the second equipment existing an interaction probability.

[0076] For example, in the embodiment, the determination of the moving area can be specifically realized based on the following method steps:

[0077] Firstly, based on the spatial view of the target space (for example, the plan layout of the factory workshop), the spatial route (that is, the path that the moving person can pass through in the industrial site) is determined, and the equipment sub-image indicating the position of each industrial equipment (that is, the identification of each industrial equipment on the plan) is identified, so that the path and equipment distribution in the industrial scene can be intuitively presented through the visual spatial view, facilitating the subsequent analysis of the relationship between the person moving and the equipment position, and ensuring that the basic layout information of the industrial environment can be accurately obtained to provide a clear spatial reference for subsequent operations;

[0078] Then, the terminal position (that is, the specific coordinates of the moving person in the industrial site) and the moving direction (that is, the direction of the moving person) of the mobile terminal are determined in the spatial view by using the mobile positioning signal (such as a mobile phone specially set with a GPS plug-in in the industrial environment) sent by the mobile terminal carried by the moving person. It can be explained that by obtaining the position and direction information of the moving person in real time, the embodiment can dynamically track the moving state of the person in the industrial space to provide real-time data support for accurately dividing the activity area, and ensure that the judgment of the moving trend of the person has timeliness and accuracy;

[0079] Then, the spatial route can be divided based on the terminal position: for example, the first divided route is opposite to the moving direction, and the second divided route is the same, so that the path can be correspondingly divided into a possible backtracking path (the first divided route) and an advancing path (the second divided route) from the moving intention of the moving person, which is more in line with the actual moving logic of the person in the industrial scene, and provides a reasonable division basis for determining the activity area of the person according to different path conditions;

[0080] Subsequently, route judgment is made based on the first divided route. If the second divided route does not exist bifurcation, the moving area of the person can be directly determined based on the second divided route, which improves the efficiency of the area determination and avoids unnecessary complex calculation, so as to quickly respond to the moving situation of the person in the industrial scene;

[0081] If the second divided route exists a branch intersection, the industrial equipment corresponding to the equipment sub-image along the first divided route is determined as the first equipment, and the industrial equipment corresponding to the equipment sub-image along the second divided route is determined as the second equipment. It can be explained that by explicitly specifying the equipment distribution on different divided routes, the interaction possibility of equipment and personnel on different paths can be analyzed, providing specific analysis objects for subsequent prediction of personnel movement area based on equipment interaction, ensuring that the determination of personnel activity area can be combined with the position distribution of actual industrial equipment.

[0082] Finally, based on the equipment interaction of the moving personnel on the first divided route (such as the interaction behaviors of the moving personnel passing by the first divided route and which industrial equipment have occurred operation, close, etc.), the interaction of each second equipment is predicted, and the personnel movement area is determined according to the second equipment with interaction probability, that is, by analyzing the equipment interaction habits of the moving personnel on the historical path (the first divided route), the interaction possibility of each equipment on the forward path (the second divided route) is predicted, so as to more accurately determine the area that the personnel may reach, fully considering the regularity and habit of personnel operation in the industrial scene, making the personnel movement area more close to the actual situation, and further accurately predicting the possible interference between personnel and industrial equipment, and improving the accuracy of safety warning in the industrial scene.

[0083] Further, in the embodiment, the above-mentioned "based on the current interaction record of the moving personnel corresponding to the first divided route, the interaction of each second equipment is predicted, and the personnel movement area corresponding to the moving personnel is determined based on the second equipment with interaction probability" can further include the following steps:

[0084] Each first equipment that is interacted by the moving personnel is input into the interaction prediction model obtained by pre-training to obtain the interaction probability corresponding to each second equipment;

[0085] In response to any second equipment having interaction probability corresponding to multiple first equipment, the interaction probability corresponding to the second equipment is fused and calculated to obtain the updated interaction probability;

[0086] The second equipment with interaction probability is determined as a preliminary interaction equipment, and a preliminary sub-route from the initial personnel position to each preliminary interaction equipment is determined based on the second divided route;

[0087] A route selection interface is created, and each preliminary sub-route arranged in descending order of corresponding interaction probability is filled into each route slot arranged in descending order in the route selection interface;

[0088] In response to the mobile terminal selecting any route slot, the personnel moving area corresponding to the mobile personnel is determined based on the preliminary sub-route filled in the route slot.

[0089] For example, in the present embodiment, the process of predicting the personnel moving area based on the interaction record can be specifically implemented based on the following method steps:

[0090] Firstly, each first device (industrial device on the first divided route (route opposite to the moving direction)) interacted by the mobile personnel on the first divided route can be input into the pre-trained interaction prediction model, so that the interaction prediction model outputs the interaction probability corresponding to each second device (industrial device on the second divided route). It can be explained that the interaction record here can be the operation of the personnel on the device, the behavior of approaching the device, etc. The interaction prediction model is constructed by learning historical interaction data, which can predict the interaction possibility of the personnel with each second device on the second divided route based on the personnel's device interaction habit on the first divided route. This step uses model prediction to convert the personnel's historical behavior into a quantitative interaction probability, so that the potential interaction relationship between the personnel and the device can be analyzed from the data level, providing data support for accurately determining the personnel moving area, and improving the scientificity and accuracy of the personnel behavior prediction;

[0091] Subsequently, when it is determined that any second device has multiple first devices corresponding to the interaction probability, the interaction probability of all the second devices needs to be fused to obtain the updated probability. For example, the interaction probability of a second device corresponding to first device A is 0.6, and the interaction probability of first device B is 0.4. In order to determine the overall interaction probability of the second device, the probabilities need to be fused. Here, the fusion calculation integrates the interaction probabilities of multiple associated devices, avoids the one-sidedness of the interaction probability of a single device, makes the prediction result more comprehensively reflect the actual interaction possibility of the personnel and the device, improves the reliability of the interaction probability, and further makes the personnel moving area determined based on the probability more consistent with the actual situation;

[0092] Then, when it is determined that any second device located on the second divided route has an interaction probability, it can be determined as a preliminary interaction device corresponding to the second device, and a preliminary sub-route from the initial personnel position to each preliminary interaction device is determined based on the second divided route (route in the moving direction). The preliminary interaction device is a device that the personnel may interact with, and the preliminary sub-route is a possible path for the personnel to reach these devices. Thus, the abstract interaction probability can be converted into a specific spatial path to determine the specific route that the personnel may travel, laying a foundation for subsequent route selection and final moving area determination, and making the determination of the personnel moving area more spatially oriented;

[0093] Then, the embodiment can create a route selection interface, and fill the preliminary sub-routes into the route slots in the interface from top to bottom in descending order of interaction probability, for example, the preliminary sub-routes with high interaction probability are arranged in the upper slots, and the preliminary sub-routes with low interaction probability are arranged in the lower slots. Here, the route selection interface presents the possible travel routes of the personnel and the corresponding interaction probabilities in a visual manner, which conforms to the use habits of the operating personnel in the industrial scene, facilitates the personnel to intuitively understand the possibilities of different routes, provides a convenient interaction mode for the personnel to autonomously select routes, and also enables more accurate determination of the movement area of the personnel based on the selection of the personnel.

[0094] Finally, when the mobile terminal selects any route slot, the embodiment can determine the movement area of the personnel based on the preliminary sub-route corresponding to the route slot. It can be explained that the actual travel route of the mobile personnel is determined through the selection of the mobile terminal, and therefore the specific activity area can be determined, and the autonomy and flexibility of the personnel operation in the industrial scene are fully considered, so that the determined movement area of the personnel is based on data prediction and conforms to the actual intention of the personnel, thereby more accurately predicting the trajectory intersection between the personnel and the industrial equipment, improving the pertinence and effectiveness of the safety warning, and ensuring that the personnel can be timely reminded to avoid the equipment operation trajectory in the industrial scene.

[0095] In addition, in the embodiment, the "fusing calculation of all interaction probabilities corresponding to the second device to obtain updated interaction probabilities" can further include the following steps:

[0096] obtaining the number of probabilities corresponding to all interaction probabilities, and determining a number evaluation value corresponding to the number dimension based on the number of probabilities;

[0097] performing mean value calculation on all interaction probabilities, and determining a mean value evaluation value corresponding to the mean value dimension based on the obtained probability;

[0098] performing weighted summation processing on the number evaluation value and the mean value evaluation value, and determining a probability growth rate based on the obtained comprehensive evaluation value;

[0099] updating the probability corresponding to the maximum probability among all interaction probabilities by the probability growth rate to obtain the updated interaction probability.

[0100] For example, in the embodiment, when the second device corresponds to the interaction probabilities of multiple first devices, the interaction probabilities need to be updated through fusion calculation, and the specific implementation manner is as follows:

[0101] Firstly, the number of all interaction probabilities is obtained, such as the interaction probability of a second device corresponding to 3 first devices, and the number of probabilities is 3. At this time, the number dimension evaluation value can be determined based on the number. Generally, the more the number, the more the interaction reference information related to the second device, and the higher the number evaluation value. For example, when the number is 3, a higher number evaluation value can be assigned according to a preset rule. Then, the potential possibility of the second device interacting with the personnel is evaluated from the data quantity by quantifying the number of probabilities, and the number evaluation value provides a reference dimension based on data richness for subsequent comprehensive evaluation, so that the fusion calculation can consider the number of first devices participating in the interaction prediction, avoid prediction deviation caused by insufficient data quantity, and improve the comprehensiveness of the interaction probability calculation;

[0102] Then, the mean value of all interaction probabilities can be calculated, such as 3 probability values being 0.5, 0.6 and 0.7, the mean value being (0.5+0.6+0.7) ÷ 3 = 0.6, and the mean value dimension mean value evaluation value is further determined based on the mean value. Here, it can be explained that the higher the mean value, the higher the overall interaction probability level, and the higher the mean value evaluation value. The mean value calculation reflects the possibility of personnel interacting with the second device from the average level dimension of the probability, and the mean value evaluation value provides a reference based on the average level of the probability for the fusion calculation, so as to consider the overall level of the interaction probability of multiple first devices, avoid the excessive influence of single high probability or low probability on the prediction result, and further improve the scientificity of the interaction probability calculation;

[0103] Then, the number evaluation value and the mean value evaluation value are weighted and summed to obtain a comprehensive evaluation value, and the probability growth rate is determined based on the comprehensive evaluation value. For example, the number evaluation value accounts for 40%, the mean value evaluation value accounts for 60%, if the number evaluation value is 80 and the mean value evaluation value is 70, the comprehensive evaluation value is 80*40%+70*60% = 74, and the corresponding probability growth rate is determined according to the size of the comprehensive evaluation value. The higher the comprehensive evaluation value, the greater the probability growth rate. It can be explained that the weighted sum processing integrates the evaluation values of the number and mean value dimensions to form a comprehensive evaluation value that comprehensively reflects the characteristics of the interaction probability, and the probability growth rate determined based on this can more reasonably reflect the comprehensive influence of the interaction probability of multiple first devices, so that the interaction probability update is more in line with the changing trend of the actual interaction possibility, and the rationality and accuracy of the interaction probability calculation are enhanced;

[0104] Finally, the maximum value in all interaction probabilities is found, such as 0.7 in the above example, and the maximum value is multiplied by the probability growth rate for updating, and then the updated interaction probability can be obtained, for example, the probability growth rate is 10%, and the updated probability is 0.7*(1+10%) = 0.77. The updating based on the maximum value highlights the main influencing factor among multiple interaction probabilities, and the probability growth rate is considered in combination with the comprehensive evaluation result, so that the updated interaction probability can not only reflect the most possible interaction, but also comprehensively consider the overall characteristics of all related interaction probabilities, so as to more accurately reflect the actual interaction possibility of the personnel and the second device, provide more reliable probability basis for subsequent determination of the personnel moving area, and then more accurately predict the interaction behavior of the personnel and the device in the industrial scene, improve the accuracy of safety warning, and ensure that the personnel can timely avoid the device running track.

[0105] It can be explained that in the present embodiment, when the interaction probability obtained through fusion calculation is greater than 1, it should be uniformly planned as 1, that is, when a certain interaction probability obtained through fusion calculation is 130%, it is planned as 100%.

[0106] In step S2, the following contents are included:

[0107] In response to the existence of the area overlapping relationship between the device moving area of any space device located in the target space and the personnel moving area, the space device is determined to have an interference attribute, and the device point cloud data of the space device with the interference attribute closest to the moving personnel and the personnel point cloud data of the moving personnel are extracted based on the obtained radar signal.

[0108] For example, in the present embodiment, the present embodiment can monitor the equipment moving area of all space equipment (industrial equipment) in the target space (such as a factory workshop, a storage area, etc.) in real time, and the personnel moving area of the moving personnel determined through the previous steps. When it is found that the equipment moving area of any space equipment and the personnel moving area exist in the area overlapping relationship, it can be determined that the space equipment has an interference attribute to the moving personnel, and then the equipment that may interfere with the personnel can be intuitively and quickly identified. For the subsequent interference analysis and early warning, it provides a clear target object, ensures that the potential safety risk source can be locked in time, improves the efficiency of safety risk identification in the industrial scene, and further, the present embodiment can extract the equipment point cloud data of the nearest equipment from the space equipment with interference attribute and the personnel point cloud data of the moving personnel based on the radar signal obtained corresponding to the target space. Here, the radar signal is obtained through the radar detection unit pre-set in the target space. The radar signal can accurately obtain the three-dimensional information such as the position and shape of the object in the target space. By extracting the equipment point cloud data of the nearest interference equipment and the personnel point cloud data, the spatial position relationship between the personnel and the equipment can be accurately described, and then it is ensured that the obtained point cloud data can accurately reflect the actual space state of the personnel and the nearest interference equipment. For the subsequent trajectory prediction and comparison, it provides high-precision data support, makes the judgment of the interference phenomenon between the personnel and the equipment more accurate, and thus improves the accuracy and reliability of safety warning in the industrial scene.

[0109] Further, in the present embodiment, the above-mentioned "in response to the area overlapping relationship between the equipment moving area of any space equipment located in the target space and the personnel moving area, determining the space equipment as having an interference attribute, and extracting the equipment point cloud data of the nearest space equipment with interference attribute corresponding to the moving personnel and the personnel point cloud data corresponding to the moving personnel based on the obtained radar signal" can further include the following steps:

[0110] creating a target twin space corresponding to the target space, and determining each equipment edge point located at the edge position of each equipment moving area and each personnel edge point located at the edge position of the personnel moving area based on the coordinate processing of the target twin space;

[0111] multiplying each equipment edge point, each personnel edge point and the retrieved preset inflation coefficient, and determining each equipment inflation model and personnel inflation model located in the target twin space based on the obtained each equipment inflation point and each personnel inflation point;

[0112] in response to the area overlapping relationship between any equipment inflation model and the personnel inflation model, determining the space equipment corresponding to the equipment inflation model as having an interference attribute;

[0113] A first distance corresponding to each device expansion model and the personnel expansion model is determined, and device point cloud data of a spatial device and personnel point cloud data of a moving personnel corresponding to the minimum first distance are extracted based on the radar detection unit.

[0114] For example, in the present embodiment, in order to accurately determine the interference relationship between the spatial device and the moving personnel and extract the point cloud data, the following technical content can be specifically used:

[0115] Firstly, a target twin space corresponding to the target space (such as a factory workshop) is created, that is, the target twin space is obtained by digitizing the actual industrial scene. Further, through coordinate processing of the target twin space, device edge points of each device moving area edge and personnel edge points of personnel moving area edge can be determined. It can be explained that the coordinate processing converts the position information of the physical space into coordinate data of the digital space, which is convenient for accurate calculation and analysis, and further converts the actual space range of the industrial device and the personnel into quantifiable edge point data, which provides an accurate space coordinate basis for subsequent expansion calculation and area overlap judgment, and ensures the accuracy of the interference attribute determination.

[0116] Then, each device edge point and personnel edge point can be multiplied by a preset expansion coefficient to obtain device expansion points and personnel expansion points, and then device expansion models and personnel expansion models are determined. Here, it can be explained that the preset expansion coefficient is set according to the requirements of the safe distance of the industrial scene, and the expansion calculation is equivalent to adding a safety buffer to the actual area of the device and the personnel respectively, fully considering the dynamic range of the industrial device running and the safety margin of the personnel activity, avoiding ignoring the potential collision risk due to simply judging based on the actual edge point, effectively improving the comprehensiveness and safety of the interference judgment, and being able to early warning possible dangerous situation;

[0117] Based on the above content, since the establishment of the device expansion model and the personnel expansion model has considered the safety buffer, when any device expansion model and the personnel expansion model have area overlap, the corresponding spatial device can be determined as having interference attribute. At this time, the overlap determination means that there is still a conflict risk between the device running and the personnel activity after considering the safety distance. This determination method combines the actual safety requirements of the industrial scene, makes the determination of the interference attribute more in line with the safety standards in the industrial environment, can accurately identify the device that really needs to be warned, reduces the false judgment and the missed judgment, and improves the practicability in the industrial scene;

[0118] Finally, the first distance (i.e., spatial distance) of each device expansion model and the personnel expansion model is calculated, and the device point cloud data of the spatial device with the minimum distance and the personnel point cloud data of the personnel are extracted by the radar detection unit. It can be explained that the radar detection unit can accurately obtain the three-dimensional point cloud information of the target, extract the device point cloud data with the minimum distance, that is, accurately locate the device with the greatest threat, provide high-priority target data for subsequent trajectory prediction, make the early warning information more targeted, and thus realize accurate early warning of safety risks in an industrial scene and ensure the safe operation of industrial production.

[0119] In step S3, the following is included:

[0120] The personnel moving trajectory obtained by the trajectory prediction of the corresponding personnel moving area based on the personnel point cloud data of the moving personnel is compared with the device moving trajectory obtained by the trajectory prediction of the corresponding device moving area based on the device point cloud data of the spatial device.

[0121] For example, in the present embodiment, based on the obtained personnel point cloud data, the present embodiment can predict the trajectory of the moving personnel in its personnel moving area to obtain a personnel moving trajectory; at the same time, based on the device point cloud data, the trajectory of the spatial device in its device moving area is predicted to obtain a device moving trajectory. It can be explained that the personnel point cloud data and the device point cloud data contain the spatial position, shape and other information of both, which provides a basis for trajectory prediction. By predicting the trajectories of the personnel and the device, the moving trend of both in a future period of time can be known in advance, so as to discover potential interference risks in time, improve the safety of personnel and device operation in an industrial scene, and by further comparing the predicted personnel moving trajectory with the device moving trajectory, it is directly shown whether the moving paths of the personnel and the device exist intersection or proximity, so as to judge whether the two will interfere at a certain moment, and thus it is convenient to discover potential safety hazards in advance, so that early warning can be given before interference occurs.

[0122] Further, in the present embodiment, the above-mentioned "comparing the personnel moving trajectory obtained by the trajectory prediction of the corresponding personnel moving area based on the personnel point cloud data of the moving personnel with the device moving trajectory obtained by the trajectory prediction of the corresponding device moving area based on the device point cloud data of the spatial device" can further include the following steps:

[0123] determining the initial personnel position, the initial personnel shape and the personnel speed of the corresponding moving personnel based on the personnel point cloud data, and determining the initial device position, the initial device shape and the device speed of the corresponding spatial device based on the device point cloud data;

[0124] obtaining a moving length of the moving area corresponding to the person, and determining a track period corresponding to the moving length based on the speed of the person, wherein the track period comprises different track moments;

[0125] determining a current position of the person corresponding to each track moment of the moving person based on the initial position of the person, and changing the initial shape of the person corresponding to each track moment based on the change rule of the shape of the person corresponding to the moving person, to obtain a current shape of the person;

[0126] determining a current position of the device corresponding to each track moment of the space device based on the initial position of the device, and changing the initial shape of the device corresponding to each track moment based on the change rule of the shape of the device corresponding to the space device, to obtain a current shape of the device;

[0127] comparing the person moving track composed of the current shape of the person and the current position of the person with the device moving track composed of the current shape of the device and the current position of the device at the same track moment.

[0128] For example, in the embodiment, the corresponding track comparison is performed based on the point cloud data, which can be specifically implemented based on the following technical contents:

[0129] Firstly, the initial position of the moving person, the initial shape of the person (for example, the contour corresponding to the standing posture of the moving person), and the speed of the person (for example, the moving speed of 0.5 m / s of the moving person) are parsed from the person point cloud data; at the same time, the initial position of the space device, the initial shape of the device (for example, the stretching state of the mechanical arm when the space device is the mechanical arm), and the speed of the device (for example, the stretching speed of the mechanical arm when the space device is the mechanical arm) are determined from the device point cloud data, so as to accurately obtain the initial motion state, provide a reliable starting point for subsequent track prediction, make the prediction result more close to the actual operation, and improve the basic accuracy of the track comparison;

[0130] Then, the moving length of the moving area corresponding to the person (for example, the total length of the path from the workstation A to the device B is 10 meters) is obtained, and the track period required to complete the length is calculated according to the speed of the person, for example, the track period is 20 seconds when the speed of the person is 0.5 m / s, and the track period is divided into different track moments (1st second, 2nd second…20th second) according to the time interval (for example, every second). It can be explained that the setting of the track period combines the actual path length and speed of the person moving in the industrial scene, so that the time range of the track prediction is more consistent with the activity period of the person in the industrial space, ensures that the time dimension of the track prediction matches the operation process of the person, avoids the error caused by too long or too short prediction time, and provides a reasonable time framework for subsequent time interval comparison;

[0131] After that, taking the initial personnel position as the starting point, the current personnel position at each time (for example, moving to 5 meters at the 5th second) is calculated according to the personnel speed and the trajectory time; at the same time, the initial personnel posture is dynamically updated according to the personnel posture change rule (for example, the limb swing model when walking), for example, the initial standing posture gradually changes to the stepping posture when walking, and the posture change at each trajectory time is preset based on the common rule of personnel movement in the industrial scene, that is, by combining the dynamic changes of position and posture, a complete moving trajectory model of personnel at different times is constructed, not only considering the position movement, but also taking into account the influence of posture change on space occupation, making the personnel trajectory prediction closer to the real scene, which can more accurately reflect the actual activity state of personnel in the industrial space, and provide a more comprehensive basis for interference judgment;

[0132] Similarly, taking the initial device position as the starting point, the current device position is calculated according to the device speed and the trajectory time (for example, the mechanical arm extends to the designated station at the 10th second), and the device posture is updated based on the device posture change rule (for example, the joint rotation model when the mechanical arm works), for example, the initial stretching state of the mechanical arm changes to the bending posture when grabbing materials, and the dynamic prediction of the device trajectory and the posture combines the running characteristics of the industrial device, such as the periodic action of the mechanical arm, so that the device trajectory model is more consistent with the actual running logic of the industrial device, so as to accurately present the spatial position and posture of the device at different times, and provide real device running data for trajectory comparison, so as to ensure that the interference judgment can accurately capture potential risks in the operation of the device;

[0133] Finally, the personnel moving trajectory (composed of the current personnel position and posture) and the device moving trajectory (composed of the current device position and posture) at the same trajectory time are compared, for example, at the 10th second, whether the posture model of the position where the personnel is located and the posture model of the current position of the device exist spatial overlap or close trend can be compared. It can be explained that through the trajectory comparison of different times and combined with the posture, the real-time interaction state of personnel and device in the industrial space can be accurately captured, not only the position is judged whether it is close, but also the space occupied by the posture of the two is considered, so that the interference judgment is more consistent with the actual risk scene of "dynamic space collision" in the industrial scene, thereby improving the accuracy and reliability of the trajectory comparison, which can discover subtle trajectory intersection hazards in advance, provide more timely and accurate decision basis for safety warning, and ensure that personnel in the industrial scene can obtain effective avoidance reminder before interference occurs.

[0134] Further, in the embodiment, the above-mentioned "comparing the personnel moving trajectory composed of the current personnel posture and the current personnel position with the device moving trajectory composed of the current device posture and the current device position at the same trajectory time" can further include the following steps:

[0135] determine a first position point and a second position point in the target twin space based on the current personnel position and the current device position at the same trajectory moment;

[0136] generate a current personnel model corresponding to the current personnel posture with the first position point as the center, and generate a current device model corresponding to the current device posture with the second position point as the center;

[0137] determine that the spatial device corresponding to the current device model and the moving personnel exist interference phenomenon in response to the model overlap relationship between the current device model and the current personnel model;

[0138] obtain a second distance between the current device model and the current personnel model in response to the non-model overlap relationship between the current device model and the current personnel model, and call the device attribute label of the spatial device corresponding to the current device model;

[0139] determine a safety distance corresponding to the spatial device based on the device attribute label, and determine that the spatial device corresponding to the current device model and the moving personnel exist interference phenomenon in response to the second distance being less than or equal to the safety distance, and vice versa.

[0140] For example, in the embodiment, whether the moving personnel and the spatial device exist interference at the same trajectory moment is accurately judged to realize the reminding of the employee avoiding the running trajectory of the industrial device, and the trajectories of the two are compared, wherein the comparison can be realized based on the following technical contents:

[0141] First, based on the current personnel position and the current device position at the same trajectory moment, a first position point and a second position point are determined in the target twin space (i.e. the space of the digital mapping of the industrial scene), for example, the target twin space can be processed by coordinate, so as to facilitate the conversion of the actual position in the industrial scene into the accurate coordinate in the digital space, and to ensure that the first position point and the second position point can accurately reflect the real-time position of the personnel and the device in the industrial scene;

[0142] Then, a current personnel model corresponding to the current personnel posture (such as the limb contour when standing or walking) is generated with the first position point as the center; at the same time, a current device model corresponding to the current device posture (such as the stretching state of the mechanical arm or the running posture of the conveyor belt) is generated with the second position point as the center, it can be explained that the generated current personnel model and the current device model can respectively intuitively represent the space occupation of the personnel and the device at the trajectory moment, and the abstract point cloud data is converted into a visual space model, which is convenient for judging the spatial relationship, fully considers the actual posture of the personnel and the device in the industrial scene, so that the interference judgment is not only based on the position, but also combines the space occupation range of the two, which improves the accuracy and comprehensiveness of the interference judgment;

[0143] Then, it is judged whether the current device model and the current personnel model have a model overlap relationship, wherein if there is an overlap, it indicates that the spatial occupation ranges of the device and the personnel have crossed at the trajectory moment, and it can be determined that the spatial device corresponding to the current device model and the moving personnel have an interference phenomenon, that is, the model overlap judgment directly reflects the collision risk that may occur between the two in the industrial space, which can quickly identify obvious interference conditions, so that timely warnings can be issued for such high-risk situations in the subsequent process to ensure personnel safety. Similarly, if the current device model and the current personnel model do not have an overlap relationship, the second distance between the two needs to be further obtained, and the device attribute label of the spatial device corresponding to the current device model is called. It can be explained that the device attribute label contains key information such as the safety distance of the device in the industrial scene. Different types of devices have different safety distances due to their operating characteristics and potential risks. For example, the safety distance of a spatial device with high temperature characteristics should be larger. By obtaining the second distance and the device attribute label, the potential risk in the non-overlapping case can be further evaluated, avoiding ignoring dangerous situations that are not overlapping but too close in distance due to only considering model overlap, making the interference judgment more comprehensive and avoiding missing any potential safety hazards.

[0144] Finally, the safety distance of the spatial device is determined based on the device attribute label. If the second distance is less than or equal to the safety distance, it indicates that although the models are not overlapping, the distance between the two has approached a dangerous range, and it is still determined that there is an interference phenomenon. Otherwise, there is no interference. By combining the safety distance judgment method, the safety requirements of different devices in the industrial scene are fully considered, making the interference judgment more in line with the actual safety standards of industrial production, and accurately identifying situations that require employees to avoid, ensuring that employees in industrial scenes can receive accurate warning information in time, effectively avoiding the operating trajectory of industrial devices, and ensuring the safe operation of industrial production.

[0145] In step S4, the following content is included:

[0146] In response to determining that the spatial device and the moving personnel have an interference phenomenon at any trajectory moment based on the comparison result, the warning information composed of the interference data generated by the trajectory moment and the corresponding interference phenomenon is sent to the mobile terminal corresponding to the moving personnel.

[0147] For example, in the embodiment, when it is determined that there is an interference phenomenon between the space equipment and the moving personnel at any trajectory moment, the mobile terminal of the moving personnel is sent a warning information, wherein the specific operation includes: first, the trajectory comparison result is monitored in real time, and when it is found that there is an interference phenomenon between the space equipment and the moving personnel at a certain trajectory moment, the trajectory moment is immediately captured, and interference data related to the interference phenomenon is generated. It can be explained that the interference data can include the specific position of the interference occurrence, the interference type (such as the intersection of the equipment running trajectory and the personnel moving path) and the severity of the interference and other information. The generation of these data is based on the position, shape and other information of the personnel and equipment obtained in the trajectory comparison process, which can accurately reflect the actual situation of the interference phenomenon. Here, timely capturing the trajectory moment and generating the interference data ensures the timeliness and accuracy of the warning information, so that the moving personnel can first understand the specific situation of the interference, and provide a reliable basis for subsequent avoidance action; then, the trajectory moment and the generated interference data can be further integrated to form a complete warning information. The integrated warning information contains the time (trajectory moment) of the interference occurrence and the specific situation (interference data) of the interference, which can comprehensively and clearly convey the potential safety risk to the moving personnel, so that the warning information is more complete and intuitive, and the moving personnel can quickly understand and respond, thereby improving the effectiveness of the warning information; finally, the warning information formed in the embodiment is sent to the mobile terminal of the corresponding moving personnel. Here, the mobile terminal can be a smart terminal device carried by the moving personnel in the industrial scene, such as an industrial tablet computer, a smart phone and the like. By sending the warning information to the mobile terminal, the moving personnel can receive the warning information in the first time, so as to have enough time to take avoidance measures to avoid interference with the running trajectory of the industrial equipment and protect their own safety. This timely and accurate warning information sending method can effectively reduce the occurrence of safety accidents in the industrial scene and improve the safety and reliability of industrial production.

[0148] Further, in the embodiment, the above-mentioned "in response to determining that there is an interference phenomenon between the space equipment and the moving personnel at any trajectory moment based on the comparison result, the warning information composed of the trajectory moment and the interference data corresponding to the interference phenomenon is sent to the mobile terminal of the corresponding moving personnel" can further include the following steps:

[0149] In response to determining that there is an interference phenomenon between any space equipment and the moving personnel at any trajectory moment based on the comparison result, the trajectory moment is determined as a starting moment, and all other trajectory moments located after the starting moment are summarized to an avoidance selection group;

[0150] generate each future device model corresponding to the current device shape at each trajectory moment in the avoidance selection group, and determine the trajectory moment closest to the starting moment as the avoidance moment, where the avoidance moment corresponds to the non-interference phenomenon between the mobile personnel and the space device;

[0151] retrieve a preset warning template, fill the device information corresponding to the space device and the avoidance period composed of the trajectory moment and the avoidance moment into the device slot and the period slot in the preset warning template, and obtain the warning information sent to the mobile terminal corresponding to the mobile personnel.

[0152] For example, in the embodiment, when it is determined based on the trajectory comparison result that the space device and the mobile personnel exist in an interference phenomenon at any trajectory moment, a warning information containing avoidance suggestions needs to be generated and sent, which can be realized based on the following technical content:

[0153] Firstly, when the comparison result shows that there is an interference phenomenon at a certain trajectory moment, the embodiment can determine this trajectory moment as the starting moment, and aggregate all trajectory moments after the starting moment into the avoidance selection group. For example, if the interference phenomenon is detected at the 10th second, the avoidance selection group contains all subsequent moments such as the 11th second, the 12th second, etc. Then, based on the time point of the interference occurrence as the starting point, the system can focus on the time range that may have risks in the future, so that the formulation of the avoidance scheme is more targeted. By specifying the starting moment and the subsequent time range, the risk time period can be locked in time in the industrial scene, and a clear time framework is provided for generating a feasible avoidance scheme, which avoids the lag of avoidance measures due to the ambiguity of the time range.

[0154] Next, the future device model corresponding to the current device shape at each trajectory moment in the avoidance selection group is generated with the second position point (i.e., the position point of the device in the target twin space) as the center. For example, the stretching state model of the device mechanical arm at the 11th second, the contraction state model at the 12th second, etc.

[0155] Subsequently, the embodiment can find in the avoidance selection group the trajectory moment closest to the starting moment where the mobile personnel and the device model do not exist in an interference phenomenon, and determine it as the avoidance moment. For example, if the device model and the personnel model do not overlap and are closest at the 15th second, then the 15th second is the avoidance moment. It can be explained that the future device model is generated in combination with the operation law of the industrial device (such as the motion cycle of the mechanical arm), so that the model can accurately predict the space state of the device at the subsequent moment. The determination of the avoidance moment gives priority to the nearest safe time point, so that the personnel can complete the avoidance in the shortest time, which meets the safety demand of “fast response” in the industrial scene. By accurately predicting the future trajectory of the device and efficiently positioning the best avoidance opportunity, the personnel are provided with clear time guidance, which improves the efficiency and safety of risk response in the industrial scene.

[0156] Finally, a pre-stored preset warning template (such as a standardized warning information format in industrial scenarios) is retrieved, and the equipment information of the spatial equipment (such as equipment name, model, location number, etc.) is filled into the equipment slot of the template. The avoidance period consisting of the start time and the avoidance time (such as "the 10th second to the 15th second") is filled into the time period slot. After generating a complete warning message, it is sent to the mobile terminal of the moving personnel. It can be said that the accurate filling of equipment information enables personnel to quickly identify risky equipment, and the clear avoidance period provides personnel with a specific time window for avoidance operations. The use of preset warning templates ensures the standardization and normalization of information format, making it easy for personnel to quickly read key information. Filling in specific data ensures the relevance and practicality of the warning content, thereby effectively reducing operational errors caused by information confusion, improving the reliability and operability of safety warnings, and ensuring that personnel can complete the avoidance in time before the equipment's operating trajectory interferes with their own movement path, thus ensuring the safe and orderly operation of industrial production.

[0157] For example, a preset warning template could include "Please note that XXX located ahead may affect personnel safety during the YYY runtime period, where 'XXX' can be understood as an equipment slot, and 'YYY' can be understood as a time slot."

[0158] Furthermore, based on the above, when interference exists at any given time along a trajectory, in order to alert personnel to avoid the interference, a warning message generated based on the interference will be sent to the personnel's mobile device. The personnel can then view the warning message on their mobile device. However, in actual industrial scenarios, since various spatial devices may be engaged in industrial production, the target space will likely experience significant ambient noise. This could cause personnel to miss the warning message on their mobile device, thus failing to recognize the potential interference and creating safety hazards. To further eliminate this drawback, this embodiment may include the following steps:

[0159] In response to the warning information being sent to the mobile terminal, the ambient light sensor unit and proximity sensor unit pre-set on the mobile terminal are identified, and a timed task for a preset duration is established.

[0160] The response is based on a timed task that determines the light collection value output by the ambient light sensing unit is less than a preset light threshold and the proximity sensing unit has not triggered a human proximity signal, and determines a warning time that is before the start time and is separated by a preset buffer period.

[0161] acquire personnel information corresponding to the moving personnel, and determine a safety number corresponding to a safety helmet worn by the moving personnel based on the personnel information;

[0162] In response to reaching the pre-warning moment, control a stimulation unit pre-set in the safety helmet corresponding to the safety number to output an electric stimulation signal.

[0163] For example, in the embodiment, in an industrial scene, to ensure that the moving personnel timely receive the avoidance reminder of the equipment running track, and avoid safety risks caused by not checking the pre-warning information of the mobile terminal, the following technical content can be further used for implementation:

[0164] First, after the pre-warning information is sent to the mobile terminal, the embodiment identifies an ambient light sensing unit (used for detecting the ambient light intensity) and a proximity sensing unit (used for sensing whether a human body is close) built in the mobile terminal, and starts a timing task lasting for a preset time length (such as 3 seconds). It can be explained that by continuously monitoring the output states of the two sensing units through the timing task, the ambient light condition of the mobile terminal and the proximity of the personnel to the mobile terminal can be acquired in real time, which provides data support for judging whether the personnel have noticed the pre-warning information, so that the situation that the personnel do not check the pre-warning due to work concentration or ambient light problems can be timely discovered, and the effectiveness of the pre-warning is ensured.

[0165] Then, if the timing task monitors that the light collection value of the ambient light sensing unit is lower than a preset light threshold value (such as 50 lux, indicating that the ambient light is relatively dark, which can be caused by the mobile terminal being placed in a pocket and not being taken out), and the proximity sensing unit does not detect a human body proximity signal (indicating that the personnel are not close to the mobile terminal), the embodiment determines a pre-warning moment located before the starting moment (the moment when the interference phenomenon occurs) and separated by a preset buffer period (such as 2 seconds). Here, the pre-warning moment before the starting moment can reserve more reaction time for the personnel, so as to avoid being unable to timely avoid due to late discovery. This pre-warning moment determination method based on the environment and the interaction state of the personnel and the equipment fully considers the situation that the personnel can not timely check the pre-warning due to the work state or environmental factors in the industrial scene, and improves the timeliness and effectiveness of the pre-warning.

[0166] Then, personnel information (such as a work number, a name, etc.) of the moving personnel is acquired, and a safety number corresponding to the safety helmet worn by the personnel is found according to the information. It can be explained that the safety number of each safety helmet is unique and can be accurately corresponded to a specific personnel, so as to ensure that the electric stimulation signal can be accurately sent to the protective equipment of the target personnel. Based on the association between the personnel information and the safety helmet number, the accurate positioning of the pre-warning measure is realized, the equipment of other personnel is avoided from being mistakenly triggered, and the accuracy and pertinence are improved.

[0167] Finally, when the preset warning time is reached, the embodiment can output an electric stimulation signal (such as a weak current pulse of 0.5 mA) through the control of the stimulation unit of the safety number safety helmet, and the electric stimulation signal directly acts on the personnel through the safety helmet, can forcibly remind the personnel in a tactile stimulation manner, and ensures that the personnel can timely receive the avoidance reminder of the device running track in the industrial scene, regardless of the working state of the personnel, effectively guarantees the safety of the personnel and the smooth progress of the industrial production.

[0168] In summary, according to the scheme of the embodiment, first, the personnel moving area is determined based on the moving trend of the moving personnel, and the device interference attribute is judged in combination with the overlapping relationship between the device moving area and the personnel moving area, which can accurately identify the industrial device that may cause interference to the employee, avoiding the misjudgment and omission problem caused by simple area division in the traditional technology; secondly, in the data processing link, the point cloud data of the device with the interference attribute closest to the moving personnel and the personnel point cloud data are extracted, which provides accurate original data support for subsequent track prediction. Based on these point cloud data, the track of the personnel and the device is predicted respectively, and the two tracks are compared, which can analyze the change trend of the personnel and the device running track in real time and dynamically. Compared with the defect that the existing technology lacks in-depth analysis of the dynamic change of the track, the method can predict the potential collision risk in advance, realize accurate early warning of dangerous situations; finally, when it is detected that the personnel and the device have interference phenomenon at a certain track moment, the warning information containing the track moment and the interference data is sent to the employee mobile terminal in time, and the employee can obtain the device running dynamic in the first time and make avoidance action in advance, which not only improves the prediction ability of the employee to the device moving trend, but also significantly reduces the safety hidden danger of the collision between the personnel and the device in the industrial scene, meets the high requirements of the industrial scene for the safe interaction between the personnel and the device, and improves the corresponding early warning accuracy.

[0169] Another embodiment of the present application provides a radar signal real-time processing system based on FPGA, Figure 2 The system includes:

[0170] The area determination module is configured to determine the area based on the moving trend of the corresponding moving personnel located in the target space to obtain the personnel moving area of the moving personnel;

[0171] The interference point cloud module is configured to determine the spatial device as having an interference attribute in response to the area overlapping relationship between the device moving area of any spatial device located in the target space and the personnel moving area, and extract the device point cloud data of the spatial device with the interference attribute closest to the moving personnel and the personnel point cloud data of the corresponding moving personnel based on the obtained radar signal;

[0172] The trajectory comparison module is configured to compare a trajectory of a moving person obtained by trajectory prediction of a corresponding moving area of the moving person based on the person point cloud data with a trajectory of a space device obtained by trajectory prediction of a corresponding moving area of the space device based on the device point cloud data.

[0173] The interference warning module is configured to, in response to determining that there is an interference phenomenon between the space device and the moving person at any trajectory moment based on the comparison result, send warning information composed of interference data generated by the trajectory moment and the corresponding interference phenomenon to a mobile terminal corresponding to the moving person.

[0174] In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the application can be practiced without these specific details. In other instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

[0175] In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the application can be practiced without these specific details. In other instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

[0176] Similarly, it is to be understood that the various features of the inventive aspects sometimes described in the specification in the context of separate embodiments can also be implemented in combination in a single embodiment. In addition, it is to be understood that the features of the embodiments sometimes described in the specification in the context of one of the inventive aspects can also be implemented and used in other ones of the inventive aspects.

[0177] Those skilled in the art understand that the modules or units or components of the devices in the examples disclosed herein can be arranged in the devices as described in the examples, or alternatively can be located in one or more devices different from the devices in the examples. The modules in the foregoing examples can be combined into one module or further divided into multiple sub-modules.

[0178] Those skilled in the art understand that the modules in the devices in the examples can be adaptively changed and arranged in one or more devices different from the examples. The modules or units or components in the examples can be combined into one module or unit or component, and further divided into multiple sub-modules or sub-units or sub-components.

[0179] Furthermore, to the extent that the terms "comprises", "comprising", "includes", "including" and "has" or any variation thereof are used in the specification and / or claims, these terms are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises, includes or has a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0180] Furthermore, some of the embodiments described herein are of a "method" or a "process" that can be embodied in software, firmware or hardware, and when embodied in software, constitutes computer-processor- executable instructions or a computer program code that can be used to program a computer processor or processors when the instructions or program code are executed. Furthermore, some of the embodiments described herein are of a "method" or a "process" that can be embodied in software, firmware or hardware, and when embodied in software, constitutes computer-processor- executable instructions or a computer program code that can be used to program a computer processor or processors when the instructions or program code are executed. Therefore, sections of the specification, including this summary and the abstract, may contain embodiments of the invention that are intended to be protected (literally and more broadly than the granted claims), but not every section contains an embodiment that is intended to be protected. Accordingly, the specification can contain protection- intended sections (which can be literally or more broadly protected) and protection-not-intended sections. The applicant(s) and / or the owner(s) of the application / family / reservation do not waive rights to protection that they would otherwise have under 35 U.S.C. § 122, et seq., merely by including sections in the specification that are not intended to be protected (or by failing to include sections intended to be protected). The specification may

[0181] As used herein, unless otherwise indicated, the use of the ordinal adjectives "first", "second", "third", etc., merely to distinguish different instances of an object to which the adjective refers, and are not intended to denote a given sequence or order of such objects. Thus, a "first" object stirred before or after a "second" object is stirred is not intended to indicate that these two objects are to be associated with a given sequence to one another, but are to be distinguished from one another only by their different associations with different instances of an object.

[0182] While the application has been described in terms of several embodiments, those skilled in the art will recognize that the application can be practiced with modifications within the spirit and scope of the application, which are intended to be encompassed by the description. Moreover, the Boolean operators "and", "or", and "not" are intended to be used in their "logical" sense, not their "Boolean" sense. That is, the use of "and" means both of the conjoined features are present, "or" means at least one of the conjoined features is present, and "not" means the absence of the feature that follows.

Claims

1. A real-time radar signal processing method based on FPGA, characterized in that, Includes the following steps: Based on the movement trend of the corresponding mobile personnel located in the target space, the region is determined to obtain the personnel movement region corresponding to the mobile personnel; If the device movement area of ​​any space device located in the target space overlaps with the personnel movement area, the space device is identified as having interference attributes. Based on the acquired radar signals, the device point cloud data of the space device with interference attributes that is closest to the moving personnel and the personnel point cloud data of the corresponding moving personnel are extracted. The trajectory of the person moving based on the trajectory prediction of the corresponding movement area of ​​the person moving based on the personnel point cloud data is compared with the trajectory prediction of the corresponding movement area of ​​the spatial device based on the device point cloud data. If the comparison results determine that there is interference between the space equipment and the mobile personnel at any trajectory time, the warning information consisting of the trajectory time and the interference data generated in response to the interference phenomenon will be sent to the mobile terminal of the corresponding mobile personnel.

2. The FPGA-based real-time radar signal processing method according to claim 1, characterized in that, Based on the movement trends of corresponding mobile personnel located in the target space, a region is determined to obtain the personnel movement region corresponding to the mobile personnel, including: Determine the spatial route and device sub-images indicating each spatial device located in the target space based on the spatial top view of the corresponding target space; Based on the mobile positioning signal sent by the mobile terminal of the corresponding mobile person, the terminal position and movement direction of the corresponding mobile terminal can be determined in the space. Based on the terminal location, the spatial route is divided into routes to obtain a first route that is opposite to the direction of movement and a second route that is in the same direction as the direction of movement. If the second route division does not have any forks, the movement area of ​​the person corresponding to the moving person is determined based on the second route division. In response to the existence of any fork in the second route, the spatial device indicated by each device sub-image along the first route is determined as the first device, and the spatial device indicated by each device sub-image along the second route is determined as the second device. Based on the device interactions of the mobile personnel corresponding to the first divided route, interaction prediction is performed for each second device, and the personnel movement area corresponding to the mobile personnel is determined based on the second devices with interaction probability.

3. The FPGA-based real-time radar signal processing method according to claim 2, characterized in that, Based on the current interaction records of the mobile personnel corresponding to the first divided route, interaction prediction is performed on each second device, and the personnel movement area corresponding to the mobile personnel is determined based on the second device corresponding to the interaction probability, including: Each first device that the mobile person will interact with is input into a pre-trained interaction prediction model to obtain the interaction probability corresponding to each second device; In response to the probability that any second device has an interaction with multiple first devices, the interaction probabilities corresponding to the second device are fused and calculated to obtain the updated interaction probabilities. The second device with a corresponding probability of interaction is identified as a potential interaction device, and a potential sub-route from the initial personnel position to each potential interaction device is determined based on the second route division. Create a route selection interface, and fill each of the prepared sub-routes, arranged from largest to smallest according to their corresponding interaction probabilities, into the corresponding route slots arranged from top to bottom in the route selection interface; In response to the mobile terminal selecting any route slot, the personnel movement area corresponding to the mobile personnel is determined based on the prepared sub-route filled in that route slot.

4. The FPGA-based real-time radar signal processing method according to claim 3, characterized in that, The updated interaction probabilities are obtained by fusing all interaction probabilities corresponding to the second device, including: Obtain the probability count for all corresponding interaction probabilities, and determine the quantity evaluation value for the corresponding quantity dimension based on the probability count; Calculate the mean of all interaction probabilities and determine the mean evaluation value of the corresponding mean dimension based on the obtained probabilities; The quantity assessment value and the mean assessment value are weighted and summed, and the probability growth rate is determined based on the resulting comprehensive assessment value. Update the probability of the interaction by increasing the corresponding probability growth rate of the one with the largest probability among all interaction probabilities, and obtain the updated interaction probability.

5. The FPGA-based real-time radar signal processing method according to claim 1, characterized in that, If the movement area of ​​any space device located in the target space overlaps with the movement area of ​​a person, the space device is identified as having interference attributes. Based on the acquired radar signals, the device point cloud data of the space device with interference attributes closest to the moving person, as well as the person point cloud data of the corresponding moving person, are extracted, including: Create a target twin space corresponding to the target space, and determine each device edge point at the edge position of each device movement area and each personnel edge point at the edge position of each personnel movement area based on the coordinate processing of the target twin space. Each device edge point and each personnel edge point is multiplied by the retrieved preset expansion coefficient, and the expansion models of each device and personnel located in the target twin space are determined based on the obtained expansion points of each device and each personnel. If any device expansion model overlaps with the personnel expansion model in a region, the spatial device corresponding to that device expansion model is identified as having interference attributes. A first distance is determined between each device expansion model and the personnel expansion model, and the device point cloud data of the spatial device with the smallest corresponding first distance and the personnel point cloud data of the moving personnel are extracted based on the radar detection unit.

6. The FPGA-based real-time radar signal processing method according to claim 5, characterized in that, The trajectory of the person, predicted based on the person's point cloud data for the corresponding movement area, is compared with the trajectory of the device, predicted based on the device's point cloud data for the corresponding movement area. This comparison includes: Based on the personnel point cloud data, the initial personnel position, initial personnel form, and personnel speed of the corresponding moving personnel are determined, and based on the equipment point cloud data, the initial equipment position, initial equipment form, and equipment speed of the corresponding spatial equipment are determined. The movement length of the corresponding personnel movement area is obtained, and the trajectory time period corresponding to the movement length is determined based on the personnel speed, wherein the trajectory time period includes different trajectory times; Based on the initial personnel position, the current personnel position of the moving personnel at each trajectory time is determined, and based on the personnel morphology change pattern of the moving personnel, the initial personnel morphology is changed at each trajectory time to obtain the current personnel morphology. Based on the initial device position, the current device position of the space device at each trajectory moment is determined, and based on the device shape change law corresponding to the space device, the initial device shape is changed at each trajectory moment to obtain the current device shape. The trajectory of personnel movement, consisting of the current personnel form and current personnel position, is compared with the trajectory of equipment movement, consisting of the current equipment form and current equipment position, at the same trajectory time.

7. The FPGA-based real-time radar signal processing method according to claim 6, characterized in that, The trajectory comparison includes: comparing the personnel movement trajectory (composed of current personnel form and current personnel position) with the equipment movement trajectory (composed of current equipment form and current equipment position) at the same trajectory time, including: Based on the current personnel position and current equipment position corresponding to the same trajectory time, the first position point and the second position point are determined in the target twin space; A current personnel model corresponding to the current personnel form is generated with the first location point as the center, and a current device model corresponding to the current device form is generated with the second location point as the center; If the current device model and the current personnel model have a model overlap relationship, it is determined that there is an interference phenomenon between the spatial equipment and the moving personnel corresponding to the current device model; In response to the fact that there is no model overlap between the current device model and the current personnel model, the second distance between the current device model and the current personnel model is obtained, and the device attribute tags of the spatial devices corresponding to the current device model are retrieved; Based on the device attribute tags, a safe distance is determined for the corresponding spatial device. If the second distance is less than or equal to the safe distance, it is determined that there is interference between the spatial device and the moving personnel corresponding to the current device model; otherwise, it is determined that there is no interference.

8. The FPGA-based real-time radar signal processing method according to claim 7, characterized in that, The response, based on comparison results, determines that interference exists between the space equipment and the mobile personnel at any given trajectory moment. It then sends a warning message, consisting of the trajectory moment and the interference data generated corresponding to the interference, to the mobile device of the corresponding mobile personnel, including: The response determines, based on the comparison results, that any spatial device and a moving person are interfering with each trajectory moment, and sets that trajectory moment as the starting moment, and aggregates all other trajectory moments after the starting moment into the avoidance selection group; Generate future device models corresponding to the current device form at each trajectory moment in the avoidance selection group, with the second location point as the center, and determine the trajectory moment that corresponds to the non-interference phenomenon with the moving person and is closest to the starting moment as the avoidance moment; A preset warning template is retrieved, and the device information corresponding to the space device, the avoidance period consisting of the trajectory time and the avoidance time are filled into the device slot and the period slot in the preset warning template to obtain the warning information sent to the mobile terminal of the corresponding mobile personnel.

9. The FPGA-based real-time radar signal processing method according to claim 8, characterized in that, The method further includes: In response to the warning information being sent to the mobile terminal, the ambient light sensor unit and proximity sensor unit pre-set on the mobile terminal are identified, and a timed task for a preset duration is established. The response is based on a timed task that determines the light collection value output by the ambient light sensing unit is less than a preset light threshold and the proximity sensing unit has not triggered a human proximity signal, and determines a warning time that is before the start time and is separated by a preset buffer period. Obtain the personnel information of the corresponding mobile personnel, and determine the safety number corresponding to the safety helmet worn by the mobile personnel based on the personnel information; Upon reaching the warning time, the system controls the stimulation unit of the safety helmet corresponding to the safety number to output an electrical stimulation signal.

10. A real-time radar signal processing system based on FPGA, characterized in that, include: The region determination module is configured to determine the region based on the movement trend of the corresponding mobile person located in the target space, and obtain the personnel movement region corresponding to the mobile person. The interference point cloud module is configured to respond to the overlap between the device movement area and the personnel movement area of ​​any space device located in the target space, identify the space device as having interference attributes, and extract the device point cloud data of the space device with interference attributes that is closest to the moving personnel and the personnel point cloud data of the corresponding moving personnel based on the acquired radar signal. The trajectory comparison module is configured to compare the trajectory of the person moving with the trajectory prediction of the corresponding person moving area based on the person point cloud data and the trajectory prediction of the corresponding device moving area based on the device point cloud data. The interference warning module is configured to respond to a situation where, based on a comparison result, it is determined that there is interference between the space device and the moving personnel at any trajectory time, and to send a warning message consisting of the trajectory time and the interference data generated in response to the interference phenomenon to the mobile terminal of the corresponding moving personnel.

Images