A method and system for preventing a ship cabin operator from being hit based on multi-sensor fusion

Abstract

This invention relates to the field of anti-collision monitoring technology for dock loading and unloading, specifically disclosing a method and system for preventing collisions with personnel working in ship holds based on multi-sensor fusion. This invention acquires the movement trajectory of the grab bucket and the movement trajectory of personnel, dynamically delineates safe and unsafe zones based on the grab bucket's movement trajectory, identifies collision risks for personnel movement trajectories in unsafe zones, acquires grab bucket movement signals and personnel movement signals based on UWB positioning units, combines the grab bucket's movement trajectory and the grab bucket's movement signals in unsafe zones to verify a first trajectory matching degree, and combines the personnel movement trajectory and the personnel movement signals to verify a second trajectory matching degree. The authenticity of the collision risk is verified based on the first and second trajectory matching degrees, and risk levels are assigned to identified collision risks. Through multi-sensor fusion, dynamic safe zone delineation, accurate risk identification, risk verification, and graded intervention are achieved, ensuring the safety of personnel working in ship holds.

Description

TECHNICAL FIELD

[0001] The present application relates to the wharf loading and unloading anti-hitting monitoring technical field, more particularly, the present application relates to a kind of ship cabin operating personnel anti-hitting method and system based on multi-sensor fusion. BACKGROUND

[0002] Due to the relatively closed and insufficiently illuminated cabin space, the cargo workers, command personnel and other personnel must enter the cabin for collaborative operation, and they often intersect with the grab bucket operation area. Due to the complex cabin environment, severe dust interference, and many visual blind areas, the traditional safety protection method relying on the experience of operating personnel and intercom communication has obvious shortcomings, and there is a risk of personnel being hit by accidental falling or swinging of the grab bucket.

[0003] Currently, single sensors such as cameras or laser radars are used to detect personnel and grab buckets, but under the conditions of low light, dust interference and cargo obstruction in the cabin, there are often problems of inaccurate recognition or high false alarm rate, and the existing systems mostly rely only on audible and visual alarm prompts, and cannot form linkage with personnel wearable devices or grab bucket mechanical control, making it difficult to block dangerous actions in time.

[0004] Therefore, it is necessary to provide a ship cabin operating personnel anti-hitting method and system based on multi-sensor fusion to solve the above technical problems, and a technical solution is provided to solve the above problems. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the present application provides a ship cabin operating personnel anti-hitting method and system based on multi-sensor fusion, which is used to solve the problem that current detection of personnel and grab buckets relies on single sensors such as cameras or laser radars, but under the conditions of low light, dust interference and cargo obstruction in the cabin, there are often problems of inaccurate recognition or high false alarm rate, and the existing systems mostly rely only on audible and visual alarm prompts, and cannot form linkage with personnel wearable devices or grab bucket mechanical control, making it difficult to block dangerous actions in time.

[0006] To achieve the above-mentioned purposes, the present application provides the following technical solutions:

[0007] A ship cabin operating personnel anti-hitting method based on multi-sensor fusion, comprising the following steps:

[0008] The laser radar is used to scan in real time to obtain the motion trajectory of the grab bucket and the motion trajectory of the personnel, and the safe area and the non-safe area are dynamically delimited according to the motion trajectory of the grab bucket;

[0009] The anti-hitting risk of the personnel motion trajectory in the non-safe area is identified, the trajectory coincidence degree is analyzed according to the motion trajectory of the grab bucket and the motion trajectory of the personnel in the non-safe area, and the anti-hitting risk is determined;

[0010] For the non-safety area where there is a risk of anti-throw, the UWB positioning unit obtains the grab movement signal and the personnel movement signal, the grab movement trajectory in the non-safety area is combined with the grab movement signal to check the first trajectory matching degree, the personnel movement trajectory is combined with the personnel movement signal to check the second trajectory matching degree, and the authenticity of the anti-throw risk is verified according to the first trajectory matching degree and the second trajectory matching degree.

[0011] For the anti-throw risk determined to exist, the risk level is divided, and the crane PLC execution module linkage grab control operation is triggered and the personnel warning is performed according to the risk level division result.

[0012] As a further scheme of the present application, the grab movement trajectory and the personnel movement trajectory are obtained by real-time scanning of the laser radar, and the safety area and the non-safety area are dynamically delimited according to the grab movement trajectory, and the specific steps are as follows:

[0013] The three-dimensional point cloud of the cabin operation area is collected by real-time scanning of the laser radar, the grab target and the personnel target are respectively identified by point cloud clustering and tracking algorithm, the movement trajectory fitting is respectively performed based on the grab target and the personnel target, the grab movement trajectory and the personnel movement trajectory are obtained, the first spatial movement feature is extracted based on the grab movement trajectory, and the second spatial movement feature is extracted based on the personnel movement trajectory; the spatial movement feature includes the speed feature and the acceleration feature;

[0014] The non-safety area is dynamically delimited by extracting the grab movement trajectory and the first spatial movement feature, and the area of the cabin operation area except the non-safety area is marked as a safety area.

[0015] As a further scheme of the present application, the non-safety area is dynamically delimited by extracting the grab movement trajectory and the first spatial movement feature, and the specific steps are as follows: the grab movement trajectory point Γ1(t)=(x1(t),y1(t),z1(t)) at time t is obtained, wherein (x1(t),y1(t),z1(t)) is the position coordinate of the grab centroid at time t, and the neighborhood Ω1(t) of the grab movement trajectory Γ1(t) at this time is defined;

[0016] The grab movement trajectory Γ1(t+ΔT) in the sampling period ΔT period is predicted;

[0017] The non-safety area is obtained based on the grab movement trajectory in the sampling period ΔT period, and Ω1(t+ΔT) is obtained; for each sampling period ΔT, the corresponding non-safety area is obtained by calculation, and the intersection operation is performed between the corresponding non-safety area and the personnel movement trajectory Γ2(t)=(x2(t),y2(t),z2(t)): R ns (t)=Ω1(t+ΔT)∩Γ2(t);

[0018] In the formula: R ns(t) is the intersection output result, Ω1(t+ΔT) is the corresponding non-safety area in the sampling period ΔT time period, Γ2(t) is the personnel motion trajectory at t moment;

[0019] According to the intersection output result, it is determined whether the personnel is in the non-safety area, if the intersection output result is a non-empty set, it is determined that the personnel is in the non-safety area, if the intersection output result is an empty set, it is determined that the personnel is in the safety area.

[0020] According to the non-safety area, the motion trajectory of the grab bucket and the personnel motion trajectory are analyzed, and the trajectory coincidence degree is analyzed, and the anti-punching risk is analyzed, and the specific steps are as follows:

[0021] If the personnel is in the non-safety area, the motion trajectory of the grab bucket and the personnel motion trajectory are obtained, and the intersection proportion of the two motion trajectories in the time window is calculated as the trajectory coincidence degree;

[0022] By comparing the trajectory coincidence degree with the preset threshold value, if the trajectory coincidence degree is greater than or equal to the preset threshold value, it is preliminarily determined that there is a risk, if the trajectory coincidence degree is less than the preset threshold value, it is preliminarily determined that there is no risk.

[0023] As a further scheme of the application, for the non-safety area with anti-punching risk, the motion signal of the grab bucket and the personnel motion signal are obtained based on the UWB positioning unit, the motion trajectory of the grab bucket in the non-safety area is combined with the motion signal of the grab bucket to check the first trajectory matching degree, the personnel motion trajectory is combined with the personnel motion signal to check the second trajectory matching degree, and the authenticity of the anti-punching risk is verified according to the first trajectory matching degree and the second trajectory matching degree, and the specific steps are as follows:

[0024] The positioning base station in the UWB positioning unit is arranged at the four corners of the cabin operation area, the first UWB signal is obtained through the first positioning tag on the personnel helmet, and the second UWB signal is obtained based on the second positioning tag on the grab bucket;

[0025] The first positioning tag and the second positioning tag are respectively measured by the four base stations, the distance from the first positioning tag to each base station is calculated as the first distance, and the distance from the second positioning tag to each base station is calculated as the second distance;

[0026] The first distance and the second distance are transmitted to the central positioning engine through the wired network;

[0027] The central positioning engine calculates the real-time three-dimensional coordinates of the grab bucket and the real-time three-dimensional coordinates of the personnel wearing the helmet according to the first distance and the second distance by using the trilateration method, the real-time three-dimensional coordinates of the grab bucket are taken as the motion signal of the grab bucket, and the real-time three-dimensional coordinates of the personnel wearing the helmet are taken as the personnel motion signal;

[0028] The movement trajectory of the grab bucket and the movement trajectory of the personnel are acquired, the Euclidean distance between the movement trajectory of the grab bucket and the movement signal of the grab bucket at the same time is calculated as a first distance value in real time, and the Euclidean distance between the movement trajectory of the personnel and the movement signal of the personnel at the same time is calculated as a second distance value; the first distance value and the second distance value are compared with a preset distance threshold value respectively, if the first distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-smashing risk is unreliable, if the first distance value is less than the preset distance threshold value, the authenticity of the anti-smashing risk is reliable, if the second distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-smashing risk is unreliable, and if the second distance value is less than the preset distance threshold value, the authenticity of the anti-smashing risk is reliable.

[0029] As a further scheme of the present application, for determining the risk level of the existing anti-smashing risk, the PLC execution module of the crane is triggered to perform linkage grab bucket control operation and personnel warning according to the risk level division result, and the specific steps are as follows:

[0030] When the authenticity of the anti-smashing risk is reliable, the safety distance is automatically adjusted according to the state of the grab bucket, and the calculation formula of the safety distance is D safe =k·(V grab ·T resp +L grab ) wherein D safe is the safety distance, k is the safety coefficient, V grab is the instantaneous speed of the grab bucket, T resp is the response time, and L grab is the projection length of the grab bucket.

[0031] According to the safety distance, the risk level is divided, and the PLC execution module of the crane is triggered to perform linkage grab bucket control operation and personnel warning according to the risk level division result.

[0032] As a further scheme of the present application, in the calculation formula of the safety distance, the safety coefficient k is 1.2 when the grab bucket is empty, and is 1.5 when the grab bucket is full.

[0033] As a further scheme of the present application, according to the risk level division result, the PLC execution module of the crane is triggered to perform linkage grab bucket control operation and personnel warning, and the specific steps include:

[0034] When D safe <D1(τ)≤1.2D safe , a first-level early warning is triggered, and at this time the employee wears a UWB bracelet for vibration reminding;

[0035] When 0.8D safe <D1(τ)≤D safe , a second-level early warning is triggered, and at this time the grab bucket is automatically decelerated, and the AR glasses worn by the personnel display the non-safety area as a red warning area;

[0036] When D1 (tau) <= 0.8D safe When D1 (tau) <= 0.8D

[0037] A ship cabin operation personnel anti-throwing system based on multi-sensor fusion comprises a perception layer, a control layer, an interaction layer and a mechanical execution layer: the perception layer comprises a laser radar, a UWB positioning unit and a grab bucket state acquisition module;

[0038] The control layer comprises a region division module, an anti-throwing risk identification module, a risk authenticity checking module and a risk grade division module;

[0039] The interaction layer comprises a UWB bracelet and AR glasses;

[0040] The mechanical execution layer comprises a crane PLC execution module.

[0041] As a further scheme of the application, the laser radar is installed at the head of the boom, and is used for real-time scanning to obtain the grab bucket motion trajectory and the personnel motion trajectory;

[0042] The UWB positioning unit comprises a positioning base station, a first positioning tag and a first positioning tag; the positioning base station is arranged at the four corners of the ship cabin operation area, the first UWB signal is obtained through the first positioning tag on the personnel helmet, and the second UWB signal is obtained based on the second positioning tag on the grab bucket; the grab bucket state acquisition module is used for obtaining the opening and closing state of the grab bucket and the load weight through PLC;

[0043] The region division module is used for real-time scanning to obtain the grab bucket motion trajectory and the personnel motion trajectory through the laser radar, and dynamically dividing the safe region and the non-safe region according to the grab bucket motion trajectory;

[0044] The anti-throwing risk identification module is used for anti-throwing risk identification on the personnel motion trajectory in the non-safe region, analyzing the trajectory coincidence degree according to the grab bucket motion trajectory and the personnel motion trajectory in the non-safe region, and identifying the anti-throwing risk;

[0045] The risk authenticity checking module is used for checking the authenticity of the anti-throwing risk in the non-safe region with the anti-throwing risk, obtaining the grab bucket motion signal and the personnel motion signal based on the UWB positioning unit, combining the grab bucket motion trajectory in the non-safe region with the grab bucket motion signal to check the first trajectory matching degree, combining the personnel motion trajectory with the personnel motion signal to check the second trajectory matching degree, and verifying the authenticity of the anti-throwing risk according to the first trajectory matching degree and the second trajectory matching degree; the risk grade division module is used for dividing the risk grade for the determined anti-throwing risk;

[0046] The UWB bracelet is used for vibration alarm to prompt personnel to evacuate;

[0047] The AR glasses are used for displaying the non-safe region as a red alert zone.

[0048] The crane PLC execution module is used to control the grab bucket operation according to the risk level classification module and to issue personnel warnings.

[0049] The technical effects and advantages of this invention, a method and system for preventing personnel from being struck by falling objects in ship cabins based on multi-sensor fusion, are as follows: This invention acquires the movement trajectory of the grab bucket and the movement trajectory of personnel, dynamically delineates safe and unsafe zones based on the grab bucket's movement trajectory, identifies the risk of falling objects by analyzing the movement trajectory of personnel in unsafe zones, acquires grab bucket movement signals and personnel movement signals based on UWB positioning units, combines the grab bucket's movement trajectory and the grab bucket's movement signals in unsafe zones to verify the first trajectory matching degree, and combines the personnel movement trajectory and the personnel movement signals to verify the second trajectory matching degree, verifies the authenticity of the risk of falling objects based on the first and second trajectory matching degrees, and classifies the risk level for the identified risk of falling objects. Through multi-sensor fusion, dynamic safe zone delineation, accurate risk identification, risk verification, and graded intervention are achieved, ensuring the safety of personnel working in ship cabins.

[0050] This invention achieves precise anti-collision protection for personnel working in the ship's cabin by multi-sensor fusion, dynamic safety zone, trajectory overlap analysis, UWB signal verification, and risk-level linkage control. It solves the problem of high false alarm rate of single sensor, enhances the reliability of the system in harsh environments, and achieves a comprehensive protection effect that combines early warning, real-time intervention, and automatic risk avoidance through human-machine linkage control mechanism. Attached Figure Description

[0051] Figure 1 A flowchart of a method for preventing personnel from being struck by objects in ship cabins based on multi-sensor fusion, provided as an embodiment of the present invention;

[0052] Figure 2 This is a system block diagram of a ship cabin worker protection system based on multi-sensor fusion, provided as an embodiment of the present invention. Detailed Implementation

[0053] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described technical solutions are only a part of this invention, and not all of it. All other technical solutions obtained by those skilled in the art based on the technical solutions of this invention without inventive effort are within the scope of protection of this invention.

[0054] like Figure 1 The diagram shown is a flowchart of a method for preventing personnel from being struck by objects in ship cabins based on multi-sensor fusion, provided by an embodiment of the present invention. Figure 1The execution subject of the method shown can be a software and / or hardware device. The execution subject of the present application can include but is not limited to at least one of the following: user equipment, network equipment, etc. Among them, the user equipment can include but is not limited to computers, smart phones, personal digital assistants (PDA) and the above-mentioned electronic devices, etc. The network equipment can include but is not limited to a single network server, a server group composed of multiple network servers, or a cloud composed of a large number of computers or network servers based on cloud computing, wherein cloud computing is a kind of distributed computing, which is a super virtual computer composed of a group of loosely coupled computers. The present embodiment does not make any limitation. Including steps S1 to S4, as follows:

[0055] S1, obtain the grab movement trajectory and the personnel movement trajectory through real-time scanning by laser radar, and dynamically delimit the safe area and the unsafe area according to the grab movement trajectory;

[0056] S2, for the personnel movement trajectory in the unsafe area, identify the anti-throwing risk, analyze the trajectory coincidence degree according to the grab movement trajectory and the personnel movement trajectory in the unsafe area, and determine the anti-throwing risk;

[0057] S3, for the unsafe area with the anti-throwing risk, obtain the grab movement signal and the personnel movement signal based on the UWB positioning unit, check the first trajectory matching degree by combining the grab movement trajectory in the unsafe area with the grab movement signal, check the second trajectory matching degree by combining the personnel movement trajectory with the personnel movement signal, and verify the authenticity of the anti-throwing risk according to the first trajectory matching degree and the second trajectory matching degree;

[0058] S4, for the determined anti-throwing risk, divide the risk level, and trigger the crane PLC execution module to perform the grab control operation and give a warning to the personnel according to the risk level division result.

[0059] Preferably, the grab movement trajectory and the personnel movement trajectory are obtained through real-time scanning by laser radar, and the safe area and the unsafe area are dynamically delimited according to the grab movement trajectory. The specific steps are as follows:

[0060] The three-dimensional point cloud of the cabin operation area is collected through real-time scanning by laser radar, the grab target and the personnel target are identified through point cloud clustering and tracking algorithm, the movement trajectory fitting is performed based on the grab target and the personnel target respectively, the grab movement trajectory and the personnel movement trajectory are obtained, the first spatial movement feature is extracted based on the grab movement trajectory, and the second spatial movement feature is extracted based on the personnel movement trajectory; the spatial movement feature includes the speed feature and the acceleration feature;

[0061] The non-safe region is dynamically delimited based on the motion trajectory of the grab bucket and the first spatial motion feature, and the region of the ship cabin operation area excluding the non-safe region is marked as a safe region.

[0062] In the ship cabin operation process, first, a laser radar device arranged at the hatch position is used to scan the operation area in real time to obtain three-dimensional point cloud information of the entire ship cabin operation environment. The point cloud data contains various target objects such as the grab bucket, the operation personnel, and the cargo. Different point cloud clusters are distinguished through a point cloud clustering algorithm, and the clustering results are dynamically identified and tracked in combination with a target tracking algorithm, so as to effectively separate and identify the grab bucket target and the personnel target. Subsequently, for the identified grab bucket target and personnel target, a fitting algorithm is used to generate the motion trajectory thereof varying with time, so as to obtain the grab bucket motion trajectory and the personnel motion trajectory.

[0063] On the basis of the trajectory generation, the first spatial motion feature of the grab bucket motion trajectory is further extracted, including the speed feature and the acceleration feature of the grab bucket in each time segment, which can accurately reflect the running state and the possible motion trend of the grab bucket. Similarly, the second spatial motion feature is extracted from the personnel motion trajectory, which is used to describe the moving speed and acceleration change of the personnel in the operation area.

[0064] Subsequently, based on the grab bucket motion trajectory and the corresponding first spatial motion feature, in combination with the running direction, the running speed, and the projection range of the grab bucket, a non-safe region, i.e., a dangerous spatial range that can be covered and affected by the grab bucket in a short time, is dynamically delimited. The non-safe region is dynamically adjusted with the real-time updating of the grab bucket motion trajectory, so as to ensure the real-time and accuracy of the dangerous region delimitation. The operation area excluding the non-safe region is automatically marked as a safe region for personnel activities and operations. In this way, dynamic safety region delimitation can be achieved in a complex ship cabin operation environment to avoid potential collision or falling risks caused by the grab bucket motion.

[0065] Preferably, the non-safe region is dynamically delimited based on the motion trajectory of the grab bucket and the first spatial motion feature, and the specific steps are as follows: a grab bucket motion trajectory point Γ1(t)=(x1(t),y1(t),z1(t)) at time t is obtained, wherein (x1(t),y1(t),z1(t)) is the position coordinate of the centroid of the grab bucket at time t, the neighborhood of the grab bucket motion trajectory at this time is defined as: Ω1(t)={(x,y,z)|||(x,y,z)-Γ1(t)||≤R(t)}; R(t)=r0+αv(t)+βa(t);

[0066] In the formula, Ω1(t) is a neighborhood of a grab bucket motion trajectory point at t time, (x, y, z) is three-dimensional coordinates of any point in the ship cabin operation area, R(t) is a neighborhood radius at t time, r0 is a geometric reference radius, alpha is a velocity gain coefficient, v(t) is a velocity characteristic in the first spatial motion characteristic, beta is an acceleration gain coefficient, and a(t) is an acceleration characteristic in the first spatial motion characteristic.

[0067] The grab bucket motion trajectory in the sampling period DeltaT time period is predicted as follows: In the formula, Γ1(t+DeltaT) is a grab bucket motion trajectory in the sampling period DeltaT time period.

[0068] The non-safe area is obtained based on the grab bucket motion trajectory prediction in the sampling period DeltaT time period, and for each sampling period DeltaT, the corresponding non-safe area is obtained by calculation, and the intersection operation is performed between the corresponding non-safe area and the personnel motion trajectory Γ2(t)=(x2(t), y2(t), z2(t)) as follows: ns (t)=Ω1(t+DeltaT)∩Γ2(t)

[0069] In the formula, R ns (t) is an intersection output result, Ω1(t+DeltaT) is a corresponding non-safe area in the sampling period DeltaT time period, and Γ2(t) is a personnel motion trajectory at t time.

[0070] The intersection output result is used to dynamically determine whether the personnel is in the non-safe area, and if the intersection output result is a non-empty set, it is determined that the personnel is in the non-safe area, and if the intersection output result is an empty set, it is determined that the personnel is in the safe area.

[0071] In the ship cabin operation process, the centroid position coordinates of the grab bucket are obtained by the laser radar at t time as Γ1(t)=(x1(t), y1(t), z1(t)). In order to evaluate the dangerous range that may be generated by the grab bucket, a dynamic neighborhood Ω1(t) needs to be established around the position point. The radius R(t) of the neighborhood is not a fixed value, but is adjusted in real time according to the geometric size r0 of the grab bucket and the motion state, for example, when the grab bucket speed v(t) is high, the neighborhood radius will increase to reflect the expansion of the dangerous area of the grab bucket at high speed, and similarly, when the grab bucket acceleration a(t) is large, the radius will also increase. The Ω1(t) defined in this way can more truly reflect the dangerous area of the grab bucket motion.

[0072] Subsequently, in a sampling period ΔT, the system predicts the next movement position Γ1(t+ΔT) of the grab bucket, and the prediction formula takes into account the velocity and acceleration of the grab bucket. For example, if the grab bucket is falling at a speed of 2 m / s and has a certain acceleration, the predicted Γ1(t+ΔT) will be closer to the true trajectory than a simple linear extrapolation in ΔT=0.5 s. Based on the predicted position, the corresponding non-safe region Ω1(t+ΔT) is generated again, thereby forming a dynamic dangerous space over time.

[0073] At the same time, the personnel movement trajectory Γ2(t) = (x2(t), y2(t), z2(t)) is also collected in real time by the positioning system. For example, a certain worker is moving along the bulkhead, and the movement trajectory points are constantly changing. The system calculates the intersection R ns (t) of Γ2(t) = (x2(t), y2(t), z2(t)) and the predicted non-safe region Ω1(t+ΔT). If the intersection result is not empty, it means that the personnel trajectory point falls into the predicted dangerous region of the grab bucket, and the system immediately determines that the personnel is in a non-safe region. For example, the calculation result shows that the personnel will enter the falling path of the grab bucket after 1 second, and the system will immediately trigger the anti-smashing risk warning. Conversely, if the intersection result is an empty set, it means that the personnel activity range has no intersection with the dangerous region of the grab bucket, and the personnel is determined to be in a safe state and no alarm is needed. The present embodiment can effectively avoid the misjudgment caused by relying on static region delimitation only, and significantly improve the accuracy and real-time performance of the anti-smashing warning, by combining the trajectory prediction of the grab bucket and the dynamic non-safe region delimitation, and the intersection judgment of the real-time trajectory of the personnel.

[0074] Preferably, the trajectory coincidence degree is analyzed according to the movement trajectory of the grab bucket in the non-safe region and the movement trajectory of the personnel, and the anti-smashing risk is determined, and the specific steps are as follows:

[0075] If the personnel is in a non-safe region, the movement trajectory of the grab bucket at this time and the movement trajectory of the personnel are obtained, the intersection ratio of the two movement trajectories in a time window is calculated as the trajectory coincidence degree, and the calculation formula is:

[0076] In the formula, η ΔT is the trajectory coincidence degree in the sampling period ΔT, D1(τ) = d(Γ1(τ), Γ2(τ)) is the Euclidean distance of the movement trajectory of the grab bucket and the movement trajectory of the personnel at time τ, d th is a dangerous threshold, δ is an indicator function, and the output result is 1 when the condition is met and 0 when the condition is not met,

[0077] is the cumulative duration of the time distance between the movement trajectory of the grab bucket and the movement trajectory of the personnel in the time window [t, t+ΔT] being less than the dangerous threshold;

[0078] By comparing the trajectory coincidence degree with the preset threshold, if the trajectory coincidence degree is greater than or equal to the preset threshold, it is preliminarily determined that there is a risk; if the trajectory coincidence degree is less than the preset threshold, it is preliminarily determined that there is no risk.

[0079] In actual ship cabin operation, when the system detects that the personnel trajectory enters the non-safe area predicted by the grab bucket, it further analyzes the coincidence of the grab bucket motion trajectory and the personnel motion trajectory in a time window. Assuming that in the time window [t, t+ΔT], the motion trajectory of the grab bucket is Γ1(τ), and the motion trajectory of the personnel is Γ2(τ), the system first calculates the Euclidean distance D1(τ)=d(Γ1(τ),Γ2(τ)) between them at each time. For example, when the grab bucket approaches the bottom of the cabin during the descending stage, and the personnel enters the hatch operation area, if the distance between them at multiple times τ is less than the dangerous threshold d th , the output of the indicator function δ(D1(τ)<d th ) is 1, indicating that there is a potential collision risk at this time; if the distance is greater than the threshold, the output is 0, indicating that this time is relatively safe.

[0080] Subsequently, the system integrates the results of all times in the entire time window [t, t+ΔT] to obtain the cumulative duration of the dangerous distance between the grab bucket and the personnel, and divides the duration by the time window length ΔT to obtain the trajectory coincidence degree η ΔT . For example, if the duration of the distance between the grab bucket and the personnel being less than the dangerous threshold is 3 seconds in a 5-second prediction time window, the trajectory coincidence degree is η ΔT =0.6.

[0081] Preferably, for the non-safe area with the risk of being hit, the grab bucket motion signal and the personnel motion signal are obtained based on the UWB positioning unit, the grab bucket motion trajectory in the non-safe area is checked with the grab bucket motion signal to obtain a first trajectory matching degree, the personnel motion trajectory is checked with the personnel motion signal to obtain a second trajectory matching degree, and the authenticity of the risk of being hit is verified according to the first trajectory matching degree and the second trajectory matching degree. The specific steps are as follows:

[0082] The positioning base stations in the UWB positioning unit are deployed at the four corners of the ship cabin operation area, the first UWB signal is obtained through the first positioning tag on the personnel helmet, and the second UWB signal is obtained based on the second positioning tag on the grab bucket;

[0083] The first positioning tag and the second positioning tag are respectively measured by the four base stations, and the distance from the first positioning tag to each base station is calculated as the first distance, and the distance from the second positioning tag to each base station is calculated as the second distance;

[0084] The first distance and the second distance are transmitted to the central positioning engine through the wired network;

[0085] The central positioning engine calculates the real-time three-dimensional coordinates of the grab bucket and the real-time three-dimensional coordinates of the person wearing the helmet according to the first distance and the second distance respectively by using the trilateration method, and takes the real-time three-dimensional coordinates of the grab bucket as the grab bucket movement signal and takes the real-time three-dimensional coordinates of the person wearing the helmet as the person movement signal.

[0086] The grab bucket movement trajectory and the person movement trajectory are obtained, and the Euclidean distance between the grab bucket movement trajectory and the grab bucket movement signal at the same moment is calculated as the first distance value, and the Euclidean distance between the person movement trajectory and the person movement signal at the same moment is calculated as the second distance value.

[0087] The first distance value and the second distance value are compared with the preset distance threshold value respectively, if the first distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-punching risk is unreliable, if the first distance value is less than the preset distance threshold value, the authenticity of the anti-punching risk is reliable, if the second distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-punching risk is unreliable, and if the second distance value is less than the preset distance threshold value, the authenticity of the anti-punching risk is reliable.

[0088] In a certain operation in the actual ship cabin operation process, the grab bucket is falling to take materials, and a worker enters the non-safe area of the grab bucket. First, the laser radar establishes the grab bucket movement trajectory and the person movement trajectory, and determines that there is a high trajectory coincidence degree between the two, so as to preliminarily determine that there is an anti-punching risk. In order to further verify the authenticity of the risk, the UWB positioning unit is started to verify.

[0089] The four UWB base stations deployed in the four corners of the ship cabin start to work. The first positioning tag installed on the worker's helmet continuously sends signals to the four base stations, and the system obtains the first distance from the tag to each base station through bidirectional ranging. Similarly, the second positioning tag installed on the grab bucket communicates with the base station to obtain the second distance from the grab bucket to each base station, and these data are transmitted to the central positioning engine in real time through a wired network.

[0090] The central positioning engine calculates the three-dimensional coordinates of the worker at this moment by using the trilateration method, and calculates the three-dimensional coordinates of the grab bucket by using the second distance. In this way, the real-time position of the worker is defined as the person movement signal, and the real-time position of the grab bucket is defined as the grab bucket movement signal.

[0091] Next, the movement trajectory of the grab bucket is compared with the real-time movement signal of the grab bucket. For example, at a certain time, the Euclidean distance between the predicted grab bucket position and the actual coordinates of the grab bucket measured by the UWB is 0.15 m, and the preset distance threshold is 0.3 m. Since 0.15 m is less than 0.3 m, it indicates that the predicted trajectory of the grab bucket is highly matched with the actual movement, and the first trajectory matching degree is reliable. Similarly, the system compares the movement trajectory of the worker with the actual position of the worker measured by the UWB. If the calculated Euclidean distance is 0.12 m, which is also less than the preset threshold 0.3 m, it indicates that the movement trajectory of the worker is highly consistent with the actual signal, and the second trajectory matching degree is reliable.

[0092] When both the first distance value and the second distance value are less than the threshold value, it is determined that the trajectory information is highly matched with the UWB signal, i.e., the authenticity of the anti-throwing risk is reliable, indicating that the worker is indeed in the dangerous area of the falling grab bucket, and there is a real anti-throwing risk, which will immediately issue a high-level warning and trigger an emergency linkage. If the predicted trajectory of the grab bucket deviates too much from the actual signal in a certain comparison, for example, the first distance value reaches 0.5 m, which is higher than the threshold value 0.3 m, it indicates that the trajectory prediction may deviate, and at this time the authenticity of the anti-throwing risk is questionable, and the system will mark the risk as unreliable and prompt to recheck.

[0093] In this way, on the basis of the preliminary risk determination, the real-time movement signal provided by the UWB positioning unit is used to calibrate the movement trajectories of the grab bucket and the worker, ensuring that the final anti-throwing risk identification has both real-time and reliability, and avoiding false positives and false negatives.

[0094] Preferably, the anti-throwing risk is divided into risk levels, and the crane PLC execution module is triggered to perform linkage grab bucket control operation and personnel warning according to the risk level division result, and the specific steps are as follows:

[0095] When the authenticity of the anti-throwing risk is reliable, the safety distance is automatically adjusted according to the state of the grab bucket, and the calculation formula of the safety distance is: safe D = k · (V grab · T resp + L grab ) In the formula, D safe is the safety distance, k is the safety coefficient, V grab is the instantaneous speed of the grab bucket, T resp is the response time, and L grab is the projection length of the grab bucket; wherein the safety coefficient k is 1.2 when the grab bucket is empty, and 1.5 when the grab bucket is full; the risk level is divided according to the safety distance, and the crane PLC execution module is triggered to perform linkage grab bucket control operation and personnel warning according to the risk level division result.

[0096] Preferably, the risk level division result triggers the crane PLC execution module to perform the linkage grab control operation and personnel warning, and the specific steps include:

[0097] When D safe <D1(τ)≤1.2D safe , a first-level early warning is triggered, and the employee wears a UWB bracelet to vibrate for reminding;

[0098] When 0.8D safe <D1(τ)≤D safe , a second-level early warning is triggered, the grab is automatically decelerated, and the AR glasses worn by the personnel display a non-safe area as a red warning area;

[0099] When D1(τ)≤0.8D safe , a third-level early warning is triggered, the grab is suspended, and an audible and visual alarm is started.

[0100] In the cabin operation of the embodiment of the application, when the anti-throwing risk is verified by the combined positioning of the laser radar and the UWB, the risk level is further divided according to the state and real-time motion parameters of the grab. At a moment, the grab is in a full load state, the system detects that the instantaneous speed of the grab is 1.5 m / s, the response time is set to 0.8 s, and the projection length of the grab is 2.0 m. First, the safety distance is calculated according to the safety distance calculation formula, wherein the safety factor k is 1.5 when full load. Substituting the data into the formula, D safe =1.5·(1.5×0.8+2.0)=4.8m, that is, the safety distance is 4.8 m.

[0101] Subsequently, the real-time relative distance D1(τ) between the grab and the worker is compared with the safety distance D safe . If it is monitored that the nearest distance between a worker and the grab is 5.2 m, D safe <D1(τ)≤1.2D safe , that is, 4.8 m < 5.2 m ≤ 5.76 m, a first-level early warning is triggered. At this time, the UWB bracelet worn by the worker vibrates to remind the personnel to pay attention to maintaining a safe distance.

[0102] If the relative distance between another worker and the grab is shortened to 4.5 m at a moment, 0.8D safe <D1(τ)≤D safe , that is, 3.84 m < 4.5 m ≤ 4.8 m, it is determined that the risk is upgraded to a second-level early warning. At this time, the grab control system automatically performs a deceleration operation, and marks the non-safe area as a red warning area in the AR glasses worn by the worker, to further strengthen the warning.

[0103] If the extreme case, a person suddenly into the grabber directly below, only with the grabber to maintain 3.5 m, meet D1(τ)≤0.8D safe , that is, ≤3.84 m, then immediately trigger a three-level early warning. At this time the grab will be suspended emergency, while triggering sound and light alarm device to ensure that the first time the workers to perceive the danger and evacuate the danger zone, so as to minimize the risk of falling accident.

[0104] A ship cabin operation personnel anti-throwing system based on multi-sensor fusion, comprising a perception layer, a control layer, an interactive layer and a mechanical execution layer:

[0105] The perception layer comprises a laser radar, a UWB positioning unit and a grab state acquisition module;

[0106] The control layer comprises a region division module, an anti-throwing risk identification module, a risk authenticity checking module and a risk level division module;

[0107] The interactive layer comprises a UWB bracelet and AR glasses;

[0108] The mechanical execution layer comprises a crane PLC execution module.

[0109] Preferably, the laser radar is installed at the head of the boom for real-time scanning to obtain the grab trajectory and the personnel trajectory; the UWB positioning unit comprises a positioning base station, a first positioning tag and a first positioning tag; the positioning base station is deployed at the four corners of the ship cabin operation area, the first UWB signal is obtained through the first positioning tag on the personnel helmet, and the second UWB signal is obtained based on the second positioning tag on the grab; the grab state acquisition module is used to obtain the opening and closing state of the grab and the load weight through the PLC;

[0110] The region division module is used to obtain the grab trajectory and the personnel trajectory through real-time scanning by the laser radar, and dynamically divide the safe area and the unsafe area according to the grab trajectory;

[0111] The anti-throwing risk identification module is used to identify the anti-throwing risk for the personnel trajectory in the unsafe area, analyze the trajectory coincidence degree according to the grab trajectory and the personnel trajectory in the unsafe area, and determine the anti-throwing risk;

[0112] The risk authenticity checking module is used to check the authenticity of the anti-throwing risk for the unsafe area with the anti-throwing risk, obtain the grab movement signal and the personnel movement signal based on the UWB positioning unit, combine the grab trajectory in the unsafe area with the grab movement signal to check the first trajectory matching degree, combine the personnel trajectory with the personnel movement signal to check the second trajectory matching degree, and verify the authenticity of the anti-throwing risk according to the first trajectory matching degree and the second trajectory matching degree; the risk level division module is used to divide the risk level for the determined anti-throwing risk;

[0113] The UWB bracelet is used for vibrating alarm to prompt personnel to evacuate;

[0114] The AR glasses are used for displaying the non-safety area as a red alert area;

[0115] The crane PLC execution module is used for controlling the operation of the grab bucket and warning personnel according to the risk level division module linkage.

[0116] As Figure 2 shown, it is a system block diagram of a ship cabin operation personnel anti-throwing system based on multi-sensor fusion, and can be used to execute the steps in the method embodiment shown, and the implementation principle and technical effects are similar, which will not be repeated here. Figure 1

[0117] Through the introduction of the above embodiments, the present application obtains the grab bucket motion trajectory and the personnel motion trajectory, dynamically divides the safety area and the non-safety area according to the grab bucket motion trajectory, identifies the anti-throwing risk for the personnel motion trajectory in the non-safety area, obtains the grab bucket motion signal and the personnel motion signal based on the UWB positioning unit, combines the grab bucket motion trajectory in the non-safety area with the grab bucket motion signal to check the first trajectory matching degree, combines the personnel motion trajectory with the personnel motion signal to check the second trajectory matching degree, verifies the authenticity of the anti-throwing risk according to the first trajectory matching degree and the second trajectory matching degree, divides the risk level for the determined anti-throwing risk, realizes dynamic safety area division, accurate risk identification, risk verification and hierarchical intervention through multi-sensor fusion, and ensures the safety of the ship cabin operation personnel.

[0118] The present application realizes accurate anti-throwing protection of the ship cabin operation personnel through multi-sensor fusion, dynamic safety area, trajectory coincidence degree analysis, UWB signal verification and risk hierarchical linkage control, solves the problem of high false alarm rate of a single sensor, strengthens the reliability of the system in harsh environments, and realizes the comprehensive protection effect of combining early warning, real-time intervention and automatic risk avoidance through the man-machine linkage control mechanism.

[0119] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be included in the protection scope of the present application.

[0120] Finally: the above is only a preferred solution of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.​

Claims

1. A method for preventing a ship cabin worker from being hit based on multi-sensor fusion, characterized in that, The method comprises the following steps: Real-time scanning by laser radar to obtain the movement trajectory of the grab bucket and the movement trajectory of the personnel, and dynamically demarcating the safe area and the non-safe area according to the movement trajectory of the grab bucket; For the movement trajectory of the personnel in the non-safe area, the anti-throwing risk identification is performed, the trajectory coincidence degree is analyzed according to the movement trajectory of the grab bucket and the movement trajectory of the personnel in the non-safe area, and the anti-throwing risk is determined; For the non-safe area with the anti-throwing risk, the movement signal of the grab bucket and the movement signal of the personnel are obtained based on the UWB positioning unit, the movement trajectory of the grab bucket in the non-safe area is checked with the movement signal of the grab bucket to obtain the first trajectory matching degree, the movement trajectory of the personnel is checked with the movement signal of the personnel to obtain the second trajectory matching degree, the authenticity of the anti-throwing risk is verified according to the first trajectory matching degree and the second trajectory matching degree, and the specific scheme comprises: Real-time calculation of the Euclidean distance between the movement trajectory of the grab bucket and the movement signal of the grab bucket at the same time as the first distance value, and calculation of the Euclidean distance between the movement trajectory of the personnel and the movement signal of the personnel at the same time as the second distance value; The first distance value and the second distance value are compared with the preset distance threshold value respectively, if the first distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-throwing risk is unreliable, if the first distance value is less than the preset distance threshold value, the authenticity of the anti-throwing risk is reliable, if the second distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-throwing risk is unreliable, and if the second distance value is less than the preset distance threshold value, the authenticity of the anti-throwing risk is reliable; For the anti-throwing risk determined to exist, the risk level is divided, and the crane PLC execution module linkage grab bucket control operation and personnel warning are triggered according to the risk level division result.

2. The method according to claim 1, wherein, Real-time scanning by laser radar to obtain the movement trajectory of the grab bucket and the movement trajectory of the personnel, and dynamically demarcating the safe area and the non-safe area according to the movement trajectory of the grab bucket, the specific steps are as follows: Real-time scanning by laser radar to obtain the movement trajectory of the grab bucket and the movement trajectory of the personnel, and dynamically demarcating the safe area and the non-safe area according to the movement trajectory of the grab bucket, the specific steps are as follows: Three-dimensional point cloud of the cabin operation area is collected by real-time scanning by laser radar, grab bucket targets and personnel targets are identified by point cloud clustering and tracking algorithm respectively, movement trajectory fitting is performed based on the grab bucket targets and the personnel targets respectively, the movement trajectory of the grab bucket and the movement trajectory of the personnel are obtained, the first spatial movement feature is extracted based on the movement trajectory of the grab bucket, and the second spatial movement feature is extracted based on the movement trajectory of the personnel; spatial The movement feature includes the speed feature and the acceleration feature; 3. The method according to claim 2, wherein the method is characterized by, The movement trajectory of the grab bucket and the first spatial movement feature are extracted to dynamically demarcate the non-safe area, and the area of the cabin operation area except the non-safe area is marked as the safe area. The movement trajectory of the grab bucket and the first spatial movement feature are extracted to dynamically demarcate the non-safe area, the specific steps are as follows: the movement trajectory point of the grab bucket at time t is obtained Γ1(t)=(x1(t),y1(t),z1(t)), wherein (x1(t), y1(t), z1(t)) is the position coordinate of the grab bucket centroid at the time, and the neighborhood Ω1(t) of the grab bucket movement trajectory Γ1(t) at this time is defined; The movement trajectory of the grab bucket in the sampling period ΔT is predicted Γ1(t+ΔT); The non-safe area is predicted based on the movement trajectory of the grab bucket in the sampling period ΔT to obtain Ω1(t+ΔT); For each sampling period ΔT, the corresponding non-safe area is obtained by calculation, and an intersection operation is performed between the corresponding non-safe area and the personnel motion trajectory Γ2(t) = (x2(t), y2(t), z2(t)): R ns (t) = Ω1(t + ΔT) ∩ Γ2(t); In the formula, R ns (t) is the intersection output result, Ω1(t+ΔT) is the corresponding non-safety area in the sampling period ΔT, and Γ2(t) is the personnel motion trajectory at the time t. The intersection output result is used to dynamically determine whether the personnel is in the non-safe area. If the intersection output result is a non-empty set, it is determined that the personnel is in the non-safe area. If the intersection output result is an empty set, it is determined that the personnel is in the safe area.

4. The method according to claim 1, wherein the method is characterized by, The trajectory coincidence degree is analyzed according to the grab motion trajectory and the personnel motion trajectory in the non-safe area, and the anti-hitting risk is determined. The specific steps are as follows: If the personnel is in the non-safe area, the grab motion trajectory and the personnel motion trajectory at this time are obtained, and the intersection proportion of the two motion trajectories in the time window is calculated as the trajectory coincidence degree. By comparing the trajectory coincidence degree with the preset threshold value, if the trajectory coincidence degree is greater than or equal to the preset threshold value, it is preliminarily determined that there is a risk. If the trajectory coincidence degree is less than the preset threshold value, it is preliminarily determined that there is no risk.

5. The method of claim 1, wherein the method is a method of preventing a punch by a cabin worker based on multi-sensor fusion. For the non-safe area with anti-hitting risk, the grab motion signal and the personnel motion signal are obtained based on the UWB positioning unit, the grab motion trajectory in the non-safe area is combined with the grab motion signal to check the first trajectory matching degree, and the personnel motion trajectory is combined with the personnel motion signal to check the second trajectory matching degree. The authenticity of the anti-hitting risk is verified according to the first trajectory matching degree and the second trajectory matching degree. The specific steps are as follows: The positioning base stations in the UWB positioning unit are arranged at the four corners of the cabin operation area. The first UWB signal is obtained through the first positioning tag on the personnel helmet, and the second UWB signal is obtained based on the second positioning tag on the grab. The first positioning tag and the second positioning tag are respectively measured by the four base stations, and the distance from the first positioning tag to each base station is calculated as the first distance, and the distance from the second positioning tag to each base station is calculated as the second distance. The first distance and the second distance are transmitted to the central positioning engine through a wired network. The central positioning engine uses the trilateration method to calculate the real-time three-dimensional coordinates of the grab and the real-time three-dimensional coordinates of the personnel wearing the helmet according to the first distance and the second distance, respectively. The real-time three-dimensional coordinates of the grab are used as the grab motion signal, and the real-time three-dimensional coordinates of the personnel wearing the helmet are used as the personnel motion signal. The grab motion trajectory and the personnel motion trajectory are obtained, the Euclidean distance between the grab motion trajectory and the grab motion signal at the same time is calculated as the first distance value, and the Euclidean distance between the personnel motion trajectory and the personnel motion signal at the same time is calculated as the second distance value. The first distance value and the second distance value are compared with the preset distance threshold value, respectively. If the first distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-hitting risk is unreliable. If the first distance value is less than the preset distance threshold value, the authenticity of the anti-hitting risk is reliable. If the second distance value is greater than or equal to the preset distance threshold value, the authenticity of the anti-hitting risk is unreliable. If the second distance value is less than the preset distance threshold value, the authenticity of the anti-hitting risk is reliable.

6. The method of claim 1, wherein the method is a method of preventing a punch by a cabin worker based on multi-sensor fusion. For determining the existence of the risk of anti-throw, according to the risk level division result, the crane PLC execution module is triggered to perform the linkage grab control operation and personnel warning, and the specific steps are as follows: When the authenticity of the anti-throw risk is reliable, the safety distance is automatically adjusted according to the grab state, and the calculation formula of the safety distance is: D safe = k - (V grab · T resp + L grab ) where: D safe V is the safety distance, k is the safety factor, V grab V is the instantaneous speed of the grab, T resp V is the response time, L grab V is the projected length of the grab; According to the risk level division result, the crane PLC execution module is triggered to perform the linkage grab control operation and personnel warning.

7. The method according to claim 6, wherein the method is characterized by, In the calculation formula of the safety distance, the safety factor k is 1.2 when the grab is empty, and 1.5 when the grab is full.

8. The method according to claim 6, wherein the method is characterized by, According to the risk level division result, the crane PLC execution module is triggered to perform the linkage grab control operation and personnel warning, and the specific steps include: When D safe <D1(τ)≤1.2D safe When this occurs, a Level 1 alert is triggered, at which point the employee wearing a UWB wristband will vibrate to alert them. When 0.8D safe When D1(τ)≤D safe When 0.8D safe When D1(τ)≤D safe When 0.8D safe When D1(τ)≤D safe When 0.8D safe When D1(τ)≤D safe When 0.8D safe When D When D1(τ)≤0.8D safe When D1(τ)≤0.8D safe When D1(τ)≤0.8D safe When D1(τ)≤0.8D safe When D1(τ)≤0.8 9. A multi-sensor fusion based anti-attack system for ship cabin workers, applied to the multi-sensor fusion based anti-attack method for ship cabin workers according to any one of claims 1-8, characterized in that, The system comprises a perception layer, a control layer, an interaction layer, and a mechanical execution layer: The perception layer comprises a laser radar, a UWB positioning unit, and a grab state acquisition module; The control layer comprises a region division module, an anti-throw risk identification module, a risk authenticity checking module, and a risk level division module; the interaction layer comprises a UWB bracelet and AR glasses; The mechanical execution layer comprises a crane PLC execution module.

10. The multi-sensor fusion based anti-ramming system for the cabin crew according to claim 9, wherein, The laser radar is installed at the head of the boom to scan and obtain the grab motion trajectory and the personnel motion trajectory in real time; The UWB positioning unit comprises a positioning base station, a first positioning tag, and a first positioning tag; the positioning base station is deployed at the four corners of the cabin operation area, the first UWB signal is obtained through the first positioning tag on the personnel helmet, and the second UWB signal is obtained based on the second positioning tag on the grab; The grab state acquisition module is used to obtain the grab opening and closing state and the load weight through the PLC; The region division module is used to obtain the grab motion trajectory and the personnel motion trajectory in real time through the laser radar, and dynamically divide the safe area and the non-safe area according to the grab motion trajectory; The anti-throw risk identification module is used to identify the anti-throw risk for the personnel motion trajectory in the non-safe area, analyze the trajectory coincidence degree based on the grab motion trajectory and the personnel motion trajectory in the non-safe area, and determine the anti-throw risk; The risk authenticity checking module is used to check the authenticity of the anti-throw risk for the non-safe area with the grab motion signal and the personnel motion signal obtained based on the UWB positioning unit, combine the grab motion trajectory in the non-safe area with the grab motion signal to check the first trajectory matching degree, combine the personnel motion trajectory with the personnel motion signal to check the second trajectory matching degree, and verify the authenticity of the anti-throw risk according to the first trajectory matching degree and the second trajectory matching degree; The risk level division module is used to divide the risk level for the determined anti-throw risk; The UWB bracelet is used to vibrate and alarm to prompt personnel to evacuate; The AR glasses are used to display the non-safe area as a red warning zone; The crane PLC execution module is used to perform the linkage grab control operation and personnel warning according to the risk level division module.

Images