A path optimization method and system for bulk cargo unloading vehicles

Abstract

The present invention relates to the technical field of path optimization for unloading vehicles, and in particular to a path optimization method and system for bulk unloading vehicles, including: obtaining the accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory; obtaining the actual time difference between the unloading vehicle to be dispatched and each traveling unloading vehicle reaching the same possible collision point; obtaining the possibility of the unloading vehicle to be dispatched and each traveling unloading vehicle colliding at the same possible collision point based on the accuracy and actual time difference of the traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory; obtaining the optimal driving path of the unloading vehicle to be dispatched based on the possibility of collision and the available paths. The present invention improves the rationality of path selection for unloading vehicles to be dispatched.

Description

Technical Field

[0001] The present invention relates to the technical field of path optimization for a discharger, and in particular to a path optimization method and system for a bulk cargo discharger. Background Art

[0002] With the development of global trade and supply chain, transportation route planning and cost control are the core competitive points in the logistics industry. The transportation route of goods directly affects the time, cost and efficiency of transportation. By rationally planning the transportation route, we can improve transportation efficiency, reduce costs and reduce resource waste.

[0003] When optimizing the path of a dump truck, the existing method selects the route with the least collision points and the shortest path as the optimized path of the dump truck. Since the time when the dump truck reaches each possible collision point on its predicted motion trajectory may be inaccurate and the path deviation may cause the path to be inaccurate, and there may be deviations in the actual time difference when the dump trucks reach the same possible collision point due to avoidance between the dump trucks, if the time accuracy, path accuracy and actual time difference are not considered, the optimized path of the dump truck will be unreasonable and a lot of resources will be wasted. Summary of the invention

[0004] In order to solve the above problems, the present invention provides a path optimization method and system for a bulk cargo unloading vehicle.

[0005] A path optimization method and system for a bulk cargo unloading vehicle of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a path optimization method for a bulk cargo unloading vehicle, the method comprising the following steps:

[0007] Obtain the remote sensing image of the port at the most recent time, obtain all the unloading vehicles in motion and one unloading vehicle to be dispatched in the port at the most recent time, and obtain the direction sequence and speed sequence of each unloading vehicle in motion in the most recent period of time;

[0008] All the moving unloading vehicles are screened out in the remote sensing image to obtain a static model image; according to the positions of the unloading vehicles to be dispatched and the unloading locations in the static model image, several optional paths of the unloading vehicles to be dispatched are obtained; according to the direction sequence and speed sequence of the moving unloading vehicles, the predicted motion trajectory of each moving unloading vehicle is obtained; according to the optional paths and the predicted motion trajectory, several possible collision points are obtained;

[0009] The direction sequence and speed sequence are fused to obtain the driving state sequence of each traveling unloading vehicle; based on the driving state sequence, the time accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained; based on the angle deviation from the starting point to the collision point on the predicted motion trajectory, the path accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained;

[0010] Combined with time accuracy and path accuracy, the accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained; the actual time difference between the unloading vehicle to be dispatched and each traveling unloading vehicle reaching the same possible collision point is obtained; based on the accuracy and actual time difference of the traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory, the possibility of a collision between the unloading vehicle to be dispatched and each traveling unloading vehicle at the same possible collision point is obtained; based on the possibility of a collision and the available paths, the optimal driving path for the unloading vehicle to be dispatched is obtained.

[0011] Furthermore, the method of obtaining a number of possible collision points according to the selected paths and the predicted motion trajectory includes the following specific steps:

[0012] The intersection points of all the alternative paths and all the predicted motion trajectories are obtained to obtain several possible collision points.

[0013] Furthermore, the direction sequence and the speed sequence are merged to obtain the driving state sequence of each driving unloading vehicle, and the specific steps include the following:

[0014] For any direction sequence and speed sequence of a moving unloading vehicle, multiply the direction values ​​and speed values ​​of the same order in the direction sequence and speed sequence to obtain a number of driving state values, and arrange all the driving state values ​​in order to form a driving state sequence.

[0015] Furthermore, the method of obtaining the time accuracy of each traveling dump truck reaching each possible collision point on its predicted motion trajectory according to the driving state sequence includes the following specific steps:

[0016] For any driving state sequence of a moving unloading vehicle; the part of the driving state sequence in the driving state sequence within the last TH1 minutes is recorded as the current driving state sequence segment, and TH1 is a preset first value; the remaining part of the driving state sequence in the driving state sequence except the current driving state sequence segment is recorded as the historical driving state sequence segment; a window of the same length as the current driving state sequence segment is obtained, and the window is moved backward along the first driving state value in the historical driving state sequence segment, one driving state value at a time, until the rightmost end of the window moves to the last driving state value in the historical driving state sequence segment, and the driving state sequence segments in the window during the movement are all recorded as reference driving state sequence segments;

[0017] For any current driving state sequence segment and several reference driving state sequence segments of a moving unloading truck, the similarity between the current driving state sequence segment and each reference driving state sequence segment is obtained, and the cumulative value of the similarity between the current driving state sequence segment and all reference driving state sequence segments and the number of reference driving state sequence segments are merged to obtain the time accuracy of the moving unloading truck reaching each possible collision point on its predicted motion trajectory; the cumulative value and the number of reference driving state sequence segments are in direct proportion to the time accuracy, and the time accuracy is a normalized value.

[0018] Furthermore, the path accuracy of each traveling unloading vehicle to each possible collision point on its predicted motion trajectory is obtained based on the angular deviation between the starting point and the collision point on the predicted motion trajectory, and the specific steps include the following:

[0019] For any possible collision point on the predicted motion trajectory of any moving unloading truck; obtain the angle deviation of each point between the starting point and the possible collision point on the predicted motion trajectory of the moving unloading truck, and the specific method for obtaining the angle deviation is: obtain the tangent direction of each point between the starting point and the possible collision point, obtain the connecting line between the starting point and the possible collision point, and use the minimum angle value between the tangent direction and the connecting line as the angle deviation of each point; integrate the angle deviation of all points from the starting point to the possible collision point, and use the inverse proportional value of the integration result as the path accuracy of the moving unloading truck to reach the possible collision point on its predicted motion trajectory.

[0020] Furthermore, the accuracy of each possible collision point of each traveling unloading vehicle on its predicted motion trajectory is obtained by combining the time accuracy and the path accuracy, and the specific steps include the following:

[0021] For any possible collision point on the predicted motion trajectory of any moving unloading truck, the time accuracy and path accuracy of the moving unloading truck arriving at the possible collision point on its predicted motion trajectory are fused to obtain the accuracy of the moving unloading truck arriving at the possible collision point on its predicted motion trajectory.

[0022] Furthermore, the specific method for obtaining the actual time difference between the unloading vehicle to be dispatched and each traveling unloading vehicle arriving at the same possible collision point is as follows:

[0023] For any moving unloading vehicle; any same possible collision point between the unloading vehicle to be started and the moving unloading vehicle is recorded as the target same possible collision point; the time consumption rating of the unloading vehicle to be started and the moving unloading vehicle to reach the target same possible collision point is obtained; the specific method for obtaining the time consumption rating is as follows: the last speed value in the speed sequence of the moving unloading vehicle is used as the speed of the moving unloading vehicle in the subsequent driving; the average of the last speed values ​​in the speed sequence of all the moving unloading vehicles is used as the speed of the unloading vehicle to be started in the subsequent driving; according to the speed of the unloading vehicle to be started and the moving unloading vehicle in the subsequent driving and the distance from the same possible collision point, the time consumption rating of the unloading vehicle to be started and the moving unloading vehicle to the target same possible collision point is obtained. The time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle and the time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle before they arrive at the target same possible collision point is obtained, and the number of time differences less than a preset second value is used as the time consumption rating for the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point; the theoretical time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point is obtained, and the theoretical time difference is the time difference obtained according to the speed of the unloading vehicle to be dispatched and the traveling unloading vehicle in subsequent driving and the distance to the target same possible collision point; the sum of the time consumption rating and the theoretical time difference is used as the actual time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point.

[0024] Furthermore, the method obtains the possibility of the unloading vehicle to be dispatched and each unloading vehicle to collide at the same possible collision point according to the accuracy and actual time difference of the unloading vehicle to be dispatched on its predicted motion trajectory; and obtains the optimal driving path of the unloading vehicle to be dispatched according to the possibility of collision and the available paths, including the following specific steps:

[0025] For any moving unloading vehicle; any common possible collision point between the unloading vehicle to be started and the moving unloading vehicle is recorded as the target common possible collision point; the actual time difference between the unloading vehicle to be started and the moving unloading vehicle reaching the target common possible collision point and the accuracy of the moving unloading vehicle reaching the possible collision point corresponding to the target common possible collision point on its predicted motion trajectory are integrated to obtain the possibility of collision between the unloading vehicle to be started and the moving unloading vehicle at the target common possible collision point;

[0026] The possible collision point corresponding to the possibility of collision greater than the preset first threshold is taken as the real collision point; the number of real collision points on each optional path of the unloading vehicle to be dispatched is obtained, and the optional path with the least number of real collision points is taken as the optimal driving path of the unloading vehicle to be dispatched.

[0027] Furthermore, the step of obtaining the predicted motion trajectory of each traveling unloading vehicle according to the direction sequence and speed sequence of the traveling unloading vehicle includes the following specific steps:

[0028] The direction sequence and speed sequence of each moving unloading truck are processed and learned through the recurrent neural network algorithm to obtain the predicted motion trajectory of each moving unloading truck.

[0029] The present invention also proposes a path optimization system for a bulk cargo unloading truck, comprising a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the steps of the aforementioned method.

[0030] The beneficial effect of the technical solution of the present invention is: after obtaining several optional paths for the unloading vehicles to be dispatched, the predicted motion trajectory of the traveling unloading vehicles and the possible collision points, the present invention obtains the time accuracy of each traveling unloading vehicle arriving at each possible collision point on its predicted motion trajectory by analyzing the driving state sequence of the traveling unloading vehicles, reducing the influence of the inaccurate time of the traveling unloading vehicles arriving at the possible collision points on the subsequent analysis of collisions, and then analyzes the angular deviation from the starting point to the collision point on the predicted motion trajectory to obtain the path accuracy of each traveling unloading vehicle arriving at each possible collision point on its predicted motion trajectory, reducing the influence of the path deviation when the traveling unloading vehicles travel along the predicted motion trajectory on the subsequent analysis of collisions, and finally obtains the possibility of the unloading vehicle to be dispatched and each traveling unloading vehicle colliding at the same possible collision point through the accuracy and actual time difference of the traveling unloading vehicle arriving at each possible collision point on its predicted motion trajectory, thereby obtaining the optimal driving path of the unloading vehicle to be dispatched, thereby improving the rationality of the path selection of the unloading vehicle to be dispatched. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 A flowchart of the steps of a path optimization method for a bulk cargo unloading vehicle provided by one embodiment of the present invention;

[0033] Figure 2 A flowchart for obtaining an optimal driving path provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the path optimization method and system of a bulk cargo unloading vehicle proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0035] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0036] The specific scheme of the path optimization method and system for a bulk cargo unloading vehicle provided by the present invention is described in detail below with reference to the accompanying drawings.

[0037] See also Figure 1 and Figure 2 , which shows a flowchart of the steps of a path optimization method for a bulk cargo unloading vehicle provided by an embodiment of the present invention and a flowchart of obtaining an optimal driving path, the method comprising the following steps:

[0038] Step S001, obtain the remote sensing image of the port at the most recent time, obtain all the unloading vehicles moving at the port at the most recent time and one unloading vehicle to be dispatched, and obtain the direction sequence and speed sequence of each unloading vehicle moving in the most recent period of time.

[0039] It should be noted that the main purpose of this embodiment is to optimize the path of the bulk cargo unloading vehicle, and the required data is collected before starting the analysis.

[0040] Specifically, a remote sensing image of the port at the most recent moment is obtained, all the unloading vehicles moving in the port at the most recent moment and one unloading vehicle to be departed are obtained, and the direction sequence and speed sequence of each unloading vehicle moving in the most recent period of time are obtained.

[0041] It should be noted that, in this embodiment, the most recent period of time is the most recent hour, the direction sequence is a sequence consisting of the movement directions of the driving unloading vehicle in the most recent period of time, and the direction sequence includes several direction values. The direction values ​​in this embodiment are angle values ​​obtained with the horizontal right direction as 0 degrees and the counterclockwise direction as the positive direction. The speed sequence is a sequence consisting of the movement speeds of the driving unloading vehicle in the most recent period of time, and the speed sequence includes several speed values. The specific acquisition method is the existing method and will not be repeated in this embodiment. The last moment of the most recent period of time is the most recent moment.

[0042] So far, the remote sensing image of the port at the most recent time, all the unloading vehicles moving in the port at the most recent time and one unloading vehicle to be dispatched, and the direction sequence and speed sequence of each unloading vehicle moving in the most recent period of time have been obtained.

[0043] Step S002, filter out all the moving unloading vehicles in the remote sensing image to obtain a static model image; obtain several optional paths for the unloading vehicles to be dispatched according to the positions of the unloading vehicles to be dispatched and the unloading locations in the static model image; obtain the predicted motion trajectory of each moving unloading vehicle according to the direction sequence and speed sequence of the moving unloading vehicles; obtain several possible collision points according to the optional paths and the predicted motion trajectories.

[0044] It should be noted that since the remote sensing image contains multiple moving unloading trucks, there is a certain interference with the path selection of the unloading truck to be dispatched. Therefore, it is necessary to obtain a static model image other than the moving unloading truck, so as to obtain several alternative paths for the unloading truck to be dispatched, and combine the predicted motion trajectory of the moving unloading truck to obtain several possible collision points, so as to facilitate the subsequent better analysis of the collision between the unloading truck to be dispatched and the moving unloading truck.

[0045] Specifically, all moving unloading vehicles are screened out in the remote sensing image to obtain a static model image. It should be noted that there are no moving objects in the static model image, which can more conveniently obtain the optional paths of the unloading vehicles to be dispatched at the port.

[0046] Furthermore, according to the positions of the unloading vehicle to be dispatched and the unloading location in the static model image, several optional paths of the unloading vehicle to be dispatched are obtained, as follows:

[0047] The unloading vehicle to be dispatched is marked in the static model image and used as the search starting point of the depth-first search algorithm. The unloading location of the unloading vehicle in the port is used as the search end point of the depth-first search algorithm in the static model image. All the optional paths of the unloading vehicle to be dispatched in the static model image are obtained through the depth-first search algorithm.

[0048] Specifically, according to the direction sequence and speed sequence of the moving unloading truck, the predicted motion trajectory of each moving unloading truck is obtained, as follows:

[0049] The direction sequence and speed sequence of each moving unloading truck are processed and learned through the Recurrent Neural Network (RNN) algorithm to obtain the predicted motion trajectory of each moving unloading truck.

[0050] It should be noted that the existing method is to process and learn the direction sequence and speed sequence of each traveling unloading truck by using a recurrent neural network (RNN) algorithm to obtain the predicted motion trajectory of each traveling unloading truck, which will not be repeated in this embodiment.

[0051] Specifically, according to the selected paths and the predicted motion trajectory, several possible collision points are obtained, as follows:

[0052] The intersection points of all the alternative paths and all the predicted motion trajectories are obtained to obtain several possible collision points.

[0053] At this point, several possible collision points are obtained.

[0054] Step S003, merge the direction sequence and the speed sequence to obtain the driving state sequence of each driving unloading vehicle; according to the driving state sequence, obtain the time accuracy of each driving unloading vehicle reaching each possible collision point on its predicted motion trajectory; according to the angle deviation from the starting point to the collision point on the predicted motion trajectory, obtain the path accuracy of each driving unloading vehicle reaching each possible collision point on its predicted motion trajectory.

[0055] It should be noted that since the time when the moving unloading truck arrives at each possible collision point on its predicted motion trajectory may be inaccurate, it will affect the subsequent analysis of the collision between the unloading truck to be dispatched and the moving unloading truck. In order to better analyze the time accuracy of the moving unloading truck arriving at each possible collision point on its predicted motion trajectory, it is necessary to fuse the direction sequence and the speed sequence to obtain the driving state sequence of each moving unloading truck. By analyzing the similarity between the current driving state sequence segment and the reference driving state sequence segment in the driving state sequence and the number of reference driving state sequence segments, the time accuracy of the moving unloading truck arriving at each possible collision point on its predicted motion trajectory can be obtained.

[0056] Specifically, the direction sequence and the speed sequence are merged to obtain the driving state sequence of each moving unloading truck. The specific method for obtaining the driving state sequence is as follows: for any direction sequence and speed sequence of a moving unloading truck, the direction values ​​and speed values ​​of the same order in the direction sequence and the speed sequence are multiplied to obtain several driving state values, and all the driving state values ​​are arranged in order to form a driving state sequence.

[0057] Specifically, according to the driving state sequence, the time accuracy of each driving unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained, including the following steps:

[0058] Firstly, according to the driving state sequence, the current driving state sequence segment and several reference driving state sequence segments of each driving unloading vehicle are obtained.

[0059] Secondly, according to the similarity between the current driving state sequence segment and the reference driving state sequence segment and the number of reference driving state sequence segments, the time accuracy of each driving unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained.

[0060] Specifically, according to the driving state sequence, the current driving state sequence segment and several reference driving state sequence segments of each driving unloading vehicle are obtained, as follows:

[0061] For any driving state sequence of a moving unloading vehicle; the part of the driving state sequence in the driving state sequence lasting TH1 minutes is recorded as the current driving state sequence segment, TH1 is a preset first value, and this embodiment is described as TH1=5; the remaining part of the driving state sequence in the driving state sequence except the current driving state sequence segment is recorded as the historical driving state sequence segment; a window with the same length as the current driving state sequence segment is obtained, and the window is moved backward along the first driving state value in the historical driving state sequence segment, one driving state value at a time, until the rightmost end of the window moves to the last driving state value in the historical driving state sequence segment, and the driving state sequence segments in the window during the moving process are all recorded as reference driving state sequence segments.

[0062] Preferably, according to the similarity between the current driving state sequence segment and the reference driving state sequence segment and the number of reference driving state sequence segments, the time accuracy of each driving dump truck reaching each possible collision point on its predicted motion trajectory is obtained, specifically as follows:

[0063] For any current driving state sequence segment and several reference driving state sequence segments of a moving unloading truck, the similarity between the current driving state sequence segment and each reference driving state sequence segment is obtained, and the cumulative value of the similarity between the current driving state sequence segment and all reference driving state sequence segments and the number of reference driving state sequence segments are merged to obtain the time accuracy of the moving unloading truck reaching each possible collision point on its predicted motion trajectory; the cumulative value and the number of reference driving state sequence segments are in direct proportion to the time accuracy, and the time accuracy is a normalized value.

[0064] As a specific example, the specific method for obtaining the time accuracy is as follows:

[0065] Any traveling unloading vehicle is recorded as a target traveling unloading vehicle; a current traveling state sequence segment of the target traveling unloading vehicle and a plurality of reference traveling state sequence segments are recorded.

[0066]

[0067] Where n is the number of reference driving state sequence segments, g iis the similarity between the current driving state sequence segment and the i-th reference driving state sequence segment. The specific method for obtaining the similarity is as follows: the average of the absolute values ​​of the differences between the current driving state sequence segment and all the driving state values ​​of the same order in the i-th reference driving state sequence segment is taken as the difference factor between the current driving state sequence segment and the i-th reference driving state sequence segment; the value of the difference factor mapped by the inverse proportional function and the normalization function is taken as the similarity between the current driving state sequence segment and the i-th reference driving state sequence segment. It should be noted that, in this embodiment, the inverse proportional relationship and normalization processing are presented through the exp{-U} model, U is the input of the model; ut is the time accuracy of the target driving unloading vehicle arriving at each possible collision point on its predicted motion trajectory, and sigmoid() is a sigmoid function used for normalization.

[0068] It should be noted that g i The larger it is, the more similar the current driving state sequence segment is to the i-th reference driving state sequence segment, and the higher the accuracy of the time when the target driving unloading vehicle arrives at each possible collision point, that is, the larger ut is; the larger n is, the more reference driving state sequence segments there are, and the longer the total duration of the reference driving state sequence segments is, that is, the more historical information is referenced, and the higher the accuracy of the time when the target driving unloading vehicle arrives at each possible collision point is.

[0069] It should be noted that since the path of the traveling unloading truck to each possible collision point on its predicted motion trajectory is also inaccurate, it will also affect the subsequent analysis of the collision between the unloading truck to be dispatched and the traveling unloading truck. Therefore, it is necessary to analyze the angular deviation on the predicted motion trajectory to obtain the accuracy of the path of the traveling unloading truck to each possible collision point on its predicted motion trajectory.

[0070] Preferably, the path accuracy of each traveling unloading vehicle to each possible collision point on its predicted motion trajectory is obtained based on the angular deviation from the starting point to the collision point on the predicted motion trajectory, as follows:

[0071] For any possible collision point on the predicted motion trajectory of any moving unloading truck; obtain the angle deviation of each point between the starting point and the possible collision point on the predicted motion trajectory of the moving unloading truck, and the specific method for obtaining the angle deviation is: obtain the tangent direction of each point between the starting point and the possible collision point, obtain the connecting line between the starting point and the possible collision point, and use the minimum angle value between the tangent direction and the connecting line as the angle deviation of each point; integrate the angle deviation of all points from the starting point to the possible collision point, and use the inverse proportional value of the integration result as the path accuracy of the moving unloading truck to reach the possible collision point on its predicted motion trajectory.

[0072] As a specific example, the specific method for obtaining the path accuracy is as follows:

[0073]

[0074] In the formula, s c is the number of points from the starting point to the cth possible collision point on the predicted motion trajectory of the target unloading vehicle; θ c,x The specific method for obtaining is as follows: obtain the tangent direction of the xth point from the starting point to the cth possible collision point, obtain the connecting line between the starting point and the cth possible collision point, and record the minimum angle between the tangent direction and the connecting line as θ c,x ; exp() is an exponential function with a natural constant as the base; l c It is the path accuracy of the target unloading truck to reach the cth possible collision point on its predicted motion trajectory.

[0075] It should be noted that θ c,x The smaller it is, the smaller the angle deviation between the starting point and the possible collision point on the predicted motion trajectory is. The smaller the deviation is, the closer the predicted motion trajectory is to a straight line, the shorter the trajectory is, and the higher the path accuracy of the target unloading truck to the cth possible collision point on its predicted motion trajectory is.

[0076] At this point, the path accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained.

[0077] Step S004, combining time accuracy and path accuracy, obtain the accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory; obtain the actual time difference between the unloading vehicle to be dispatched and each traveling unloading vehicle reaching the same possible collision point; based on the accuracy and actual time difference of the traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory, obtain the possibility of the unloading vehicle to be dispatched and each traveling unloading vehicle colliding at the same possible collision point; based on the possibility of collision and the available paths, obtain the optimal driving path for the unloading vehicle to be dispatched.

[0078] It should be noted that since the time for the moving unloading truck to reach each possible collision point on its predicted motion trajectory may be inaccurate, and the path for the moving unloading truck to reach each possible collision point on its predicted motion trajectory may also be inaccurate, the above analysis analyzes the time accuracy and path accuracy, and combines the two to comprehensively obtain the accuracy of the moving unloading truck to reach each possible collision point on its predicted motion trajectory.

[0079] Preferably, the accuracy of each possible collision point of each traveling unloading vehicle on its predicted motion trajectory is obtained by combining the time accuracy and the path accuracy, as follows:

[0080] For any possible collision point on the predicted motion trajectory of any moving unloading truck, the time accuracy and path accuracy of the moving unloading truck arriving at the possible collision point on its predicted motion trajectory are fused to obtain the accuracy of the moving unloading truck arriving at the possible collision point on its predicted motion trajectory.

[0081] As a specific example, a specific method for obtaining the accuracy of the traveling unloading vehicle reaching the possible collision point on its predicted motion trajectory is as follows:

[0082] f c =ut c × c

[0083] In the formula, ut c is the time accuracy of the target unloading vehicle reaching the cth possible collision point on its predicted motion trajectory; c is the path accuracy of the target unloading vehicle reaching the cth possible collision point on its predicted motion trajectory; f c It is the accuracy of the target unloading vehicle reaching the cth possible collision point on its predicted motion trajectory.

[0084] It should be noted that ut c The larger the value is, the greater the accuracy of the time when the target unloading vehicle reaches the cth possible collision point on its predicted motion trajectory; l c The larger it is, the greater the accuracy of the path of the target unloading vehicle reaching the cth possible collision point on its predicted motion trajectory, and the greater the accuracy of the target unloading vehicle reaching the cth possible collision point on its predicted motion trajectory.

[0085] It should be noted that when the time for the unloading truck to be dispatched and the moving unloading truck to arrive at the same possible collision point is relatively close, avoidance may occur, which will affect the time difference between the unloading truck to be dispatched and the moving unloading truck arriving at the same possible collision point. Therefore, by analyzing the time taken for the unloading truck to be dispatched and the moving unloading truck to arrive at the same possible collision point, the actual time difference between the unloading truck to be dispatched and the moving unloading truck to arrive at the same possible collision point is obtained.

[0086] Preferably, the actual time difference between the unloading vehicle to be dispatched and each traveling unloading vehicle arriving at the same possible collision point is obtained, specifically as follows:

[0087] For any moving unloading vehicle; any same possible collision point between the unloading vehicle to be started and the moving unloading vehicle is recorded as the target same possible collision point; the time consumption rating of the unloading vehicle to be started and the moving unloading vehicle to reach the target same possible collision point is obtained; the specific method for obtaining the time consumption rating is as follows: the last speed value in the speed sequence of the moving unloading vehicle is used as the speed of the moving unloading vehicle in the subsequent driving; the average of the last speed values ​​in the speed sequence of all the moving unloading vehicles is used as the speed of the unloading vehicle to be started in the subsequent driving; according to the speed of the unloading vehicle to be started and the moving unloading vehicle in the subsequent driving and the distance from the same possible collision point, the time consumption rating of the unloading vehicle to be started and the moving unloading vehicle to the target same possible collision point is obtained. The time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle and the time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle before they arrive at the target same possible collision point is obtained, and the number of time differences less than a preset second value is used as the time consumption rating for the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point; the theoretical time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point is obtained, and the theoretical time difference is the time difference obtained according to the speed of the unloading vehicle to be dispatched and the traveling unloading vehicle in subsequent driving and the distance to the target same possible collision point; the sum of the time consumption rating and the theoretical time difference is used as the actual time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point.

[0088] As a specific example, the specific method for obtaining the actual time difference is as follows:

[0089] B h =k h +b h

[0090] In the formula, k h is a time rating for the unloading vehicle to be started and the target unloading vehicle to reach the hth same possible collision point. The specific method for obtaining the time rating is as follows: the last speed value in the speed sequence of the target unloading vehicle is used as the speed of the target unloading vehicle in the subsequent driving; the average of the last speed values ​​in the speed sequence of all the unloading vehicles is used as the speed of the unloading vehicle to be started in the subsequent driving; according to the speeds of the unloading vehicle to be started and the target unloading vehicle in the subsequent driving and the distances from the same possible collision point, the time difference between the unloading vehicle to be started and the target unloading vehicle to reach the same possible collision point is obtained; the time difference of all the same possible collision points before the unloading vehicle to be started and the target unloading vehicle reach the hth same possible collision point is obtained, and the number of time differences less than the preset second value is used as the time rating for the unloading vehicle to be started and the target unloading vehicle to reach the hth same possible collision point. This embodiment is described with the second value of 1 minute. h B is the theoretical time difference between the unloading vehicle to be dispatched and the target unloading vehicle arriving at the same possible collision point h; his the actual time difference between the unloading truck to be dispatched and the target unloading truck arriving at the hth possible collision point.

[0091] It should be noted that, since there will be avoidance when the time for the unloading truck to be dispatched and the target unloading truck to reach the same possible collision point is less than 1 minute, a certain amount of time may be consumed before reaching the hth same possible collision point, resulting in inaccurate time difference obtained based on distance and speed. Therefore, the time rating of the unloading truck to be dispatched and the target unloading truck to reach the hth same possible collision point is analyzed, k h The larger it is, the more avoidance situations there are, and the actual time difference will increase. By combining the theoretical time difference, the actual time difference between the unloading vehicle to be dispatched and the target unloading vehicle arriving at the same possible h-th collision point is obtained. The theoretical time difference is the time difference calculated without considering the avoidance situation.

[0092] It should be noted that the above analysis includes the accuracy of the moving unloading truck reaching the possible collision point on its predicted motion trajectory, and the actual time difference between the unloading truck to be dispatched and each moving unloading truck reaching the same possible collision point. Both of them can reflect the possibility of a collision between the unloading truck to be dispatched and each moving unloading truck at the same possible collision point. Therefore, by analyzing the accuracy and actual time difference of the moving unloading truck reaching each possible collision point on its predicted motion trajectory, the possibility of a collision between the unloading truck to be dispatched and each moving unloading truck at the same possible collision point is obtained.

[0093] Preferably, the probability of the unloading vehicle to be dispatched colliding with each unloading vehicle at the same possible collision point is obtained based on the accuracy and actual time difference of the unloading vehicle arriving at each possible collision point on its predicted motion trajectory, as follows:

[0094] For any moving unloading truck, any same possible collision point between the unloading truck to be dispatched and the moving unloading truck is recorded as the target same possible collision point. The actual time difference between the unloading truck to be dispatched and the moving unloading truck in reaching the target same possible collision point and the accuracy of the moving unloading truck reaching the possible collision point corresponding to the target same possible collision point on its predicted motion trajectory are integrated to obtain the possibility of a collision between the unloading truck to be dispatched and the moving unloading truck at the target same possible collision point.

[0095] As a specific example, the specific method for obtaining the possibility of collision occurring at the same possible collision point is as follows:

[0096] ε c =exp(-B c )×f c

[0097] In the formula, B cis the actual time difference between the unloading vehicle to be dispatched and the target unloading vehicle arriving at the same possible collision point c; exp() is an exponential function with a natural constant as the base; f c is the accuracy of the target unloading vehicle reaching the cth possible collision point on its predicted motion trajectory; ε c is the possibility of collision between the unloading truck to be started and the target unloading truck at the same possible collision point c.

[0098] It should be noted that B c The smaller it is, the smaller the actual time difference between the unloading vehicle to be started and the target unloading vehicle arriving at the cth possible collision point is, the closer the time between the unloading vehicle to be started and the target unloading vehicle arriving at the cth possible collision point is, and the greater the possibility of collision at the cth possible collision point is; at the same time, f c The larger it is, the higher the accuracy of the target unloading truck reaching the cth possible collision point on its predicted motion trajectory, and the greater the possibility of collision at the cth same possible collision point.

[0099] It should be noted that the above-obtained possibility of collision between the unloading vehicle to be dispatched and each traveling unloading vehicle at the same possible collision point is optimized by selecting the path with the least collision points on the available paths of the unloading vehicle to be dispatched.

[0100] Furthermore, based on the possibility of collision and the available paths, the optimal driving path of the unloading vehicle to be dispatched is obtained, as follows:

[0101] The possible collision point corresponding to the preset first threshold value for the possibility of collision is taken as the real collision point; the number of real collision points on each optional path of the unloading vehicle to be dispatched is obtained, and the optional path with the least number of real collision points is taken as the optimal driving path of the unloading vehicle to be dispatched. This embodiment is described with the first threshold value of 0.65.

[0102] Through the above steps, a path optimization method for bulk cargo unloading vehicles is completed.

[0103] Another embodiment of the present invention provides a path optimization system for a bulk cargo unloading vehicle, the system comprising a memory and a processor, and when the processor executes a computer program stored in the memory, the processor performs the following operations:

[0104] Acquire the remote sensing image of the port at the most recent moment, acquire all the unloading vehicles in motion and one unloading vehicle to be dispatched at the most recent moment, and acquire the direction sequence and speed sequence of each unloading vehicle in motion in the most recent period; screen out all unloading vehicles in motion from the remote sensing image to obtain a static model image; obtain several alternative paths for the unloading vehicles to be dispatched according to the positions of the unloading vehicles to be dispatched and the unloading locations in the static model image; obtain the predicted motion trajectory of each unloading vehicle according to the direction sequence and speed sequence of the unloading vehicles; obtain several possible collision points according to the alternative paths and the predicted motion trajectory; fuse the direction sequence and the speed sequence to obtain the driving state sequence of each unloading vehicle in motion; obtain the time required for each unloading vehicle to reach its predicted motion state according to the driving state sequence. The time accuracy of each possible collision point on the moving trajectory; according to the angle deviation from the starting point to the collision point on the predicted motion trajectory, the path accuracy of each traveling unloading vehicle to reach each possible collision point on its predicted motion trajectory is obtained; combining the time accuracy and path accuracy, the accuracy of each traveling unloading vehicle to reach each possible collision point on its predicted motion trajectory is obtained; the actual time difference between the unloading vehicle to be departed and each traveling unloading vehicle arriving at the same possible collision point is obtained; according to the accuracy and actual time difference of the traveling unloading vehicle to reach each possible collision point on its predicted motion trajectory, the possibility of the unloading vehicle to be departed and each traveling unloading vehicle colliding at the same possible collision point is obtained; based on the possibility of collision and the available paths, the optimal driving path of the unloading vehicle to be departed is obtained.

[0105] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A path optimization method for a bulk cargo unloading vehicle, characterized in that: The method comprises the following steps: Obtain the remote sensing image of the port at the most recent time, obtain all the unloading vehicles in motion and one unloading vehicle to be dispatched in the port at the most recent time, and obtain the direction sequence and speed sequence of each unloading vehicle in motion in the most recent period of time; All the moving unloading vehicles are screened out in the remote sensing image to obtain a static model image; according to the positions of the unloading vehicles to be dispatched and the unloading locations in the static model image, several optional paths of the unloading vehicles to be dispatched are obtained; according to the direction sequence and speed sequence of the moving unloading vehicles, the predicted motion trajectory of each moving unloading vehicle is obtained; according to the optional paths and the predicted motion trajectory, several possible collision points are obtained; The direction sequence and speed sequence are fused to obtain the driving state sequence of each traveling unloading vehicle; based on the driving state sequence, the time accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained; based on the angle deviation from the starting point to the collision point on the predicted motion trajectory, the path accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained; Combined with the time accuracy and path accuracy, the accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory is obtained; the actual time difference between the unloading vehicle to be dispatched and each traveling unloading vehicle reaching the same possible collision point is obtained; based on the accuracy and actual time difference of the unloading vehicle to be dispatched and each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory, the possibility of the unloading vehicle to be dispatched and each traveling unloading vehicle colliding at the same possible collision point is obtained; based on the possibility of collision and the available paths, the optimal driving path of the unloading vehicle to be dispatched is obtained; The specific steps of obtaining the time accuracy of each traveling unloading vehicle reaching each possible collision point on its predicted motion trajectory according to the driving state sequence are as follows: For any driving state sequence of a moving unloading vehicle; The partial driving state sequence of minutes is recorded as the current driving state sequence segment. is a preset first value; the remaining driving state sequence in the driving state sequence except the current driving state sequence segment is recorded as a historical driving state sequence segment; a window of the same length as the current driving state sequence segment is obtained, and the window is moved backward along the first driving state value in the historical driving state sequence segment, one driving state value at a time, until the rightmost end of the window moves to the last driving state value in the historical driving state sequence segment, and the driving state sequence segments in the window during the movement are all recorded as reference driving state sequence segments; For any current driving state sequence segment of a driving unloading vehicle and several reference driving state sequence segments; obtain the similarity between the current driving state sequence segment and each reference driving state sequence segment, merge the accumulated value of the similarity between the current driving state sequence segment and all reference driving state sequence segments and the number of reference driving state sequence segments, and obtain the time accuracy of the driving unloading vehicle reaching each possible collision point on its predicted motion trajectory; the accumulated value, the number of reference driving state sequence segments and the time accuracy are proportional, and the time accuracy is a normalized value; The specific method for obtaining the time accuracy is as follows: Any traveling unloading vehicle is recorded as a target traveling unloading vehicle; for the current driving state sequence segment of the target traveling unloading vehicle and several reference driving state sequence segments; In the formula, is the number of reference driving state sequence segments, The current driving state sequence segment and the The specific method for obtaining the similarity is: comparing the current driving state sequence segment with the first reference driving state sequence segment. The average of the absolute values ​​of the differences between all driving state values ​​of the same order in the first reference driving state sequence segment is taken as the difference between the current driving state sequence segment and the first reference driving state sequence segment. The difference factor of the reference driving state sequence segment is used as the value after the difference factor is mapped by the inverse proportional function and the normalization function as the difference between the current driving state sequence segment and the first driving state sequence segment. Similarity of reference driving state sequence segments; The time accuracy of the target unloading vehicle reaching each possible collision point on its predicted motion trajectory, is the sigmoid function; The specific steps of obtaining the path accuracy of each traveling unloading vehicle to each possible collision point on its predicted motion trajectory according to the angular deviation between the starting point and the collision point on the predicted motion trajectory are as follows: For any possible collision point on the predicted motion trajectory of any moving unloading vehicle; obtain the angle deviation of each point between the starting point and the possible collision point on the predicted motion trajectory of the moving unloading vehicle, and the specific method for obtaining the angle deviation is: obtain the tangent direction of each point between the starting point and the possible collision point, obtain the connecting line between the starting point and the possible collision point, and use the minimum angle value between the tangent direction and the connecting line as the angle deviation of each point; integrate the angle deviation of all points between the starting point and the possible collision point, and use the inverse proportional value of the integration result as the path accuracy of the moving unloading vehicle to reach the possible collision point on its predicted motion trajectory; The specific method for obtaining the path accuracy is as follows: In the formula, The predicted motion trajectory of the target unloading vehicle from the starting point to the The number of points between possible collision points; The specific method of obtaining is as follows: Get the number of The possible collision points The tangent direction of the point is obtained by The connection line between the possible collision points, the minimum angle between the tangent direction and the connection line is recorded as ; is an exponential function with a natural constant as base; The target unloading vehicle reaches the first position on its predicted motion trajectory. The path accuracy of each possible collision point.

2. A path optimization method for a bulk cargo unloading vehicle according to claim 1, characterized in that: The specific steps of obtaining several possible collision points according to the selected paths and predicted motion trajectories are as follows: The intersection points of all the alternative paths and all the predicted motion trajectories are obtained to obtain several possible collision points.

3. The path optimization method for a bulk cargo unloading vehicle according to claim 1, characterized in that: The direction sequence and the speed sequence are merged to obtain the driving state sequence of each driving unloading vehicle, and the specific steps include the following: For any direction sequence and speed sequence of a moving unloading vehicle, multiply the direction values ​​and speed values ​​of the same order in the direction sequence and speed sequence to obtain a number of driving state values, and arrange all the driving state values ​​in order to form a driving state sequence.

4. The path optimization method for a bulk cargo unloading vehicle according to claim 1, characterized in that: The method combines the time accuracy and the path accuracy to obtain the accuracy of each possible collision point of each traveling unloading vehicle on its predicted motion trajectory, including the following specific steps: For any possible collision point on the predicted motion trajectory of any moving unloading truck, the time accuracy and path accuracy of the moving unloading truck arriving at the possible collision point on its predicted motion trajectory are fused to obtain the accuracy of the moving unloading truck arriving at the possible collision point on its predicted motion trajectory.

5. The path optimization method for a bulk cargo unloading vehicle according to claim 1, characterized in that: The specific method for obtaining the actual time difference between the unloading vehicle to be dispatched and each unloading vehicle in motion reaching the same possible collision point is as follows: For any moving unloading vehicle; any same possible collision point between the unloading vehicle to be started and the moving unloading vehicle is recorded as the target same possible collision point; the time consumption rating of the unloading vehicle to be started and the moving unloading vehicle to reach the target same possible collision point is obtained; the specific method for obtaining the time consumption rating is as follows: the last speed value in the speed sequence of the moving unloading vehicle is used as the speed of the moving unloading vehicle in the subsequent driving; the average of the last speed values ​​in the speed sequence of all the moving unloading vehicles is used as the speed of the unloading vehicle to be started in the subsequent driving; according to the speed of the unloading vehicle to be started and the moving unloading vehicle in the subsequent driving and the distance from the same possible collision point, the time consumption rating of the unloading vehicle to be started and the moving unloading vehicle to the target same possible collision point is obtained. The time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle and the time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle before they arrive at the target same possible collision point is obtained, and the number of time differences less than a preset second value is used as the time consumption rating for the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point; the theoretical time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point is obtained, and the theoretical time difference is the time difference obtained according to the speed of the unloading vehicle to be dispatched and the traveling unloading vehicle in subsequent driving and the distance to the target same possible collision point; the sum of the time consumption rating and the theoretical time difference is used as the actual time difference between the unloading vehicle to be dispatched and the traveling unloading vehicle to arrive at the target same possible collision point.

6. A path optimization method for a bulk cargo unloading vehicle according to claim 1, characterized in that: The method obtains the possibility of the unloading vehicle to be dispatched and each unloading vehicle to collide at the same possible collision point according to the accuracy and actual time difference of the unloading vehicle to be dispatched on its predicted motion trajectory; obtains the optimal driving path of the unloading vehicle to be dispatched according to the possibility of collision and the available paths, including the following specific steps: For any moving unloading vehicle; any common possible collision point between the unloading vehicle to be started and the moving unloading vehicle is recorded as the target common possible collision point; the actual time difference between the unloading vehicle to be started and the moving unloading vehicle reaching the target common possible collision point and the accuracy of the moving unloading vehicle reaching the possible collision point corresponding to the target common possible collision point on its predicted motion trajectory are integrated to obtain the possibility of collision between the unloading vehicle to be started and the moving unloading vehicle at the target common possible collision point; The possible collision point corresponding to the collision possibility greater than the preset first threshold is regarded as the real collision point; The number of real collision points on each optional path of the unloading vehicle to be dispatched is obtained, and the optional path with the least number of real collision points is used as the optimal driving path of the unloading vehicle to be dispatched.

7. The path optimization method for a bulk cargo unloading vehicle according to claim 1, characterized in that: The specific steps of obtaining the predicted motion trajectory of each traveling unloading vehicle according to the direction sequence and speed sequence of the traveling unloading vehicle are as follows: The direction sequence and speed sequence of each moving unloading truck are processed and learned through the recurrent neural network algorithm to obtain the predicted motion trajectory of each moving unloading truck.

8. A path optimization system for a bulk cargo unloading vehicle, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by a processor, the steps of a path optimization method for a bulk cargo unloading truck as described in any one of claims 1-7 are implemented.

Images