A method and device for processing monitoring data from multiple sources
By performing time synchronization and trajectory fusion in digital twin monitoring technology, the problem of signal data ejaculation and delay is solved, and the monitoring accuracy and effect are improved.
Patent Information
- Application Number
- CN202411178599.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-27
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-27
AI Technical Summary
在应用数字孪生监控技术时,目标的信号数据飘逸、网络传输延迟和处理耗时导致监控目标的时间精度和位置精度不高,监控效果较差。
By obtaining the monitoring data processed by each positioning source server, time synchronization is performed using designated coordinated world time, multiple local trajectories are identified and fused to obtain tracking trajectories, and the extended trajectory is estimated based on the monitoring data, supplementing the motion state of the delayed time process, and improving monitoring time accuracy and position accuracy.
By processing the tracking trajectory and extended trajectory, the time accuracy and position accuracy of monitoring are improved, thereby improving the monitoring effect.
Smart Images

Figure CN118709144B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a method and device for processing monitoring data from multiple sources. Background Art
[0002] With the continuous development of digital twin technology, digital twin monitoring technology is widely used in all walks of life. Digital twin monitoring technology is used to monitor the trajectories of various targets, such as using digital twin monitoring technology to monitor the flight path of aircraft.
[0003] However, in the process of using digital twin monitoring technology to monitor various targets, due to problems such as target signal data fluctuations, network transmission delays, and time-consuming processing, this will lead to low time accuracy and position accuracy of the monitored targets, and poor monitoring effects. Summary of the invention
[0004] In view of this, an embodiment of the present invention provides a method and device for processing monitoring data from multiple sources to solve the problem of poor monitoring effect when using digital twin monitoring technology to monitor various targets.
[0005] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:
[0006] A first aspect of an embodiment of the present invention discloses a method for processing monitoring data from multiple sources, the method comprising:
[0007] Acquire monitoring data processed by each positioning source server, each positioning source server adopts a designated coordinated universal time for time synchronization, and the monitoring data at least includes a local trajectory of the monitoring target;
[0008] From the monitoring data processed by each of the positioning source servers, a plurality of local trajectories of a specific object are identified, the specific object being any of the monitoring targets;
[0009] fusing a plurality of local trajectories of the specific object to obtain a tracking trajectory of the specific object;
[0010] Based on the monitoring data processed by each of the positioning source servers, an extended trajectory of the specific object is estimated.
[0011] Preferably, identifying a plurality of local trajectories of a specific object from the monitoring data processed by each of the positioning source servers includes:
[0012] Using the monitoring data processed by each of the positioning source servers, construct a trajectory set;
[0013] Classifying the trajectory set to obtain at least one classification cluster, each of which contains all local trajectories of the same monitoring target;
[0014] Any of the classification clusters is obtained to obtain a plurality of local trajectories of a specific object.
[0015] Preferably, fusing the multiple local trajectories of the specific object to obtain the tracking trajectory of the specific object includes:
[0016] The weight coefficients corresponding to the respective positioning source servers are used to perform weighted summation on the multiple local trajectories of the specific object to obtain the tracking trajectory of the specific object.
[0017] Preferably, estimating the extended trajectory of the specific object based on the monitoring data processed by each of the positioning source servers includes:
[0018] estimating at least one candidate trajectory of the specific object based on the monitoring data processed by each of the positioning source servers;
[0019] Evaluating collision risk and logical cost of the candidate trajectory;
[0020] Establishing a search space containing all the candidate trajectories;
[0021] Based on the collision risk and the logical cost, the candidate trajectory that meets a specific condition is searched from the search space to obtain an extended trajectory of the specific object.
[0022] Preferably, it also includes:
[0023] The tracking trajectory and the extended trajectory are uploaded in sequence as a reference trajectory of the specific object.
[0024] A second aspect of an embodiment of the present invention discloses a device for processing monitoring data from multiple sources, the device comprising:
[0025] An acquisition unit, configured to acquire monitoring data processed by each positioning source server, each positioning source server adopting a designated coordinated universal time for time synchronization, and the monitoring data at least including a local track of the monitoring target;
[0026] an identification unit, configured to identify a plurality of local tracks of a specific object from the monitoring data processed by each of the positioning source servers, wherein the specific object is any monitoring target;
[0027] A fusion unit, used for fusing multiple local trajectories of the specific object to obtain a tracking trajectory of the specific object;
[0028] The estimating unit is used to estimate the extended trajectory of the specific object based on the monitoring data processed by each of the positioning source servers.
[0029] Preferably, the identification unit comprises:
[0030] A construction subunit, configured to construct a trajectory set using the monitoring data processed by each of the positioning source servers;
[0031] A classification subunit, used for classifying the trajectory set to obtain at least one classification cluster, each of which contains all local trajectories of the same monitoring target;
[0032] The acquisition subunit is used to acquire any of the classification clusters to obtain multiple local trajectories of a specific object.
[0033] Preferably, the fusion unit is specifically used to: perform weighted summation on multiple local trajectories of the specific object using weight coefficients corresponding to each of the positioning source servers to obtain the tracking trajectory of the specific object.
[0034] Preferably, the estimation unit comprises:
[0035] an estimating subunit, configured to estimate at least one candidate trajectory of the specific object based on the monitoring data processed by each of the positioning source servers;
[0036] An evaluation subunit, used for evaluating the collision risk and logical cost of the candidate trajectory;
[0037] Establishing a subunit, used to establish a search space containing all the candidate trajectories;
[0038] The search subunit is configured to search the candidate trajectory satisfying a specific condition from the search space based on the collision risk and the logical cost, so as to obtain an extended trajectory of the specific object.
[0039] Preferably, it also includes:
[0040] The uploading unit is used to sequentially upload the tracking trajectory and the extended trajectory as a reference trajectory of the specific object.
[0041] Based on the above-mentioned embodiment of the present invention, a method and device for processing monitoring data from multiple sources are provided. The method is as follows: obtaining monitoring data processed by each positioning source server, each positioning source server adopts a specified coordinated universal time for time synchronization; identifying multiple local trajectories of a specific object from the monitoring data processed by each positioning source server, the specific object being any monitoring target; fusing multiple local trajectories of the specific object to obtain a tracking trajectory of the specific object; estimating an extended trajectory of the specific object based on the monitoring data processed by each positioning source server. In this solution, monitoring data processed by each positioning source server that adopts a specified coordinated universal time for time synchronization is obtained, and multiple local trajectories of the specific object are identified. Multiple local trajectories of the specific object are fused to obtain a tracking trajectory of the specific object, and an extended trajectory of the specific object is estimated. The tracking trajectory and the extended trajectory are processed to supplement the motion state of the delayed time process, so as to improve the time accuracy and position accuracy of the monitoring, thereby improving the monitoring effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] 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 embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.
[0043] Figure 1 An example diagram of a physical space and a virtual space provided by an embodiment of the present invention;
[0044] Figure 2 An example diagram of a delay time model provided by an embodiment of the present invention;
[0045] Figure 3 A flowchart of a method for processing monitoring data from multiple sources provided by an embodiment of the present invention;
[0046] Figure 4 A schematic diagram of the principle of the extrapolation and interpolation method provided by an embodiment of the present invention;
[0047] Figure 5 A diagram showing an example of the weighted summation principle provided by an embodiment of the present invention;
[0048] Figure 6 An example diagram of the principle of obtaining the tracking trajectory of a specific object by fusion provided in an embodiment of the present invention;
[0049] Figure 7 A flowchart of estimating an extended trajectory of a specific object provided by an embodiment of the present invention;
[0050] Figure 8An example diagram of decomposing trajectory data and environmental data into a grid provided by an embodiment of the present invention;
[0051] Fig. 9 An example diagram of the overall process of a method for processing monitoring data from multiple sources provided by an embodiment of the present invention;
[0052] Fig.10 A structural block diagram of a device for processing monitoring data from multiple sources provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0054] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.
[0055] With the continuous development of digital twin technology, digital twin monitoring technology is widely used in all walks of life. Technologies such as twin operation monitoring and remote control have replaced traditional application models. However, in the process of using digital twin monitoring technology to monitor various targets, there will be problems such as target signal data fluctuations, network transmission delays, and time-consuming processing. This will result in low time accuracy and position accuracy of the monitored targets, and poor monitoring effects.
[0056] For example: traditional optical camera monitoring has problems such as incomplete monitoring information and small monitoring viewing angle. Although the use of digital twin monitoring technology can improve digital monitoring capabilities, due to problems such as target signal data fluctuations, network transmission delays, and time-consuming processing, the time accuracy and position accuracy of the monitored target will be low.
[0057] Therefore, this solution proposes a method and device for processing monitoring data from multiple sources, obtaining monitoring data processed by each positioning source server that uses a specified coordinated universal time for time synchronization, and identifying multiple local trajectories of a specific object. The multiple local trajectories of the specific object are fused to obtain the tracking trajectory of the specific object, and the extended trajectory of the specific object is estimated. The tracking trajectory and the extended trajectory are processed to supplement the motion state of the delayed time process, so as to improve the time accuracy and position accuracy of the monitoring, thereby improving the monitoring effect.
[0058] More specifically, this solution constructs a multi-source data fusion method to perform target detection, data association, tracking and identification, situation assessment and prediction on the monitoring data of multiple sensors and the same monitoring target, and uses real-time data enhancement to improve the continuity and accuracy of the virtual and real states under different communication and interaction delay interferences, thereby improving the time accuracy and location accuracy of monitoring, and thus improving the monitoring effect.
[0059] Through research, it is found that in order to solve the problem of inconsistent accuracy of multi-source data, various target positioning technologies are different in positioning accuracy, coverage, update cycle, positioning method, etc. For example, the specific contents of some positioning technologies are shown in Table 1.
[0060] Table 1:
[0061]
[0062] To solve the problem of inconsistent synchronization accuracy between virtual and real space, the operation of the entire monitoring system is divided into physical space and virtual space, such as Figure 1 As can be seen from the example diagram of the physical space and the virtual space shown, according to the change of time, the physical space and the virtual space are divided into a sending phase (first phase), a receiving phase (second phase) and a suspend request phase (third phase);
[0063] Due to the time delay in data transmission, synchronization anomalies occur in the first stage of physical space and the second stage of virtual space; the third stage can be predicted to be consistent through virtual-real synchronization.
[0064] In response to the above problems, this solution improves the continuity and accuracy of monitoring by constructing a delay time model and fusing trajectories. Here, the delay time model is explained.
[0065] The delay problem is caused by network transmission and monitoring system interaction time. A delay time model is pre-built. The specific situation of the delay time model is as follows: Figure 2 shown.
[0066] Figure 2In the figure, the horizontal axis (X axis) represents the different states of the data message from the physical space to the virtual space, and the vertical axis (Y axis) represents the time relationship between different states; T tgt is the target state time, T est+fus is the estimated motion time, T cur The current time when the data is received, T send is the physical space transmission time, ε is the network delay, △T ce is the time difference from the current state to the motion estimation state;
[0067] △T ct is the time difference from the current state to the target state, ξ send is the physical space sending state, ξ cur is the virtual space data receiving state, ξ est+fus is the motion estimation state, ξ tgt is the target state;
[0068] Figure 2 There are three cases in the send ->T cur ", "From T cur ->T est+fus ", "From T cur ->T tgt "; Among them, from ξ cur To est+fus The reference trajectory is called TT (which is the tracking trajectory mentioned later). est+fus To tgt The reference trajectory is called ET (also referred to as the extended trajectory later).
[0069] It should be noted that the monitoring targets mentioned in this plan can also be referred to as "targets" for short.
[0070] See also Figure 3 , shows a flow chart of a method for processing monitoring data from multiple sources provided by an embodiment of the present invention, the method comprising:
[0071] Step S301: Acquire monitoring data processed by each positioning source server.
[0072] It should be noted that each positioning source server uses a specified coordinated universal time for time synchronization. For example, each positioning source server uses GPS coordinated universal time for time synchronization. The monitoring data processed by each positioning source server contains at least a local trajectory of the monitored target (at least a local trajectory of at least one monitored target). The following first explains the method of time synchronization of the positioning source server.
[0073] The positioning device (also called the positioning source) sends the collected positioning data to the positioning source server (also called the positioning source processor). The positioning source server tracks the monitoring target by processing the positioning data, thereby obtaining the corresponding monitoring data.
[0074] It should be noted that since each positioning source is independently timed, each positioning source does not have a unified time reference, each positioning source does not start detection at the same time, and the relative position of each positioning source to the monitored target is different; in the above case, each positioning source will not obtain positioning data at the same time, that is, there will be differences in the detection data time of each positioning source.
[0075] Therefore, time synchronization (also called time alignment) is required before trajectory processing is performed on the positioning data transmitted by each positioning source. For example, GPS timing can be used as a unified time reference.
[0076] Specifically, the precise clock signal of GPS is sent to each positioning source server through the fusion center, so that the clocks of all positioning sources use GPS Coordinated Universal Time.
[0077] The monitoring data output by each positioning source server is timestamped, so that the fusion center can clearly know the acquisition time of the monitoring data, which makes it easier to align the monitoring data of the same monitoring target to a unified time. Specifically, after all positioning sources use GPS Coordinated Universal Time, each positioning source server uses the same time reference. At this time, the aforementioned "aligning the monitoring data of the same monitoring target to a unified time" can be achieved by extrapolation and interpolation based on the time when each positioning source acquires the positioning data of the monitoring target.
[0078] For example: through Figure 4 The extrapolation and interpolation method shown achieves the aforementioned "aligning the monitoring data of the same monitoring target to the same time"; Figure 4 In the above example, the start detection time is the same, T(k-1) to T(k) is a detection cycle of the positioning sensor (positioning device), T ij (i=1,2,…,n) respectively represent the detection time of positioning sensor 1 to positioning sensor n for the jth monitoring target.
[0079] Through the above method, the monitoring data of the same monitoring target are aligned to the same time.
[0080] In the specific implementation of step S301, monitoring data processed by each positioning source server is obtained, and the monitoring data processed by each positioning source server contains detailed information of a local track of at least one monitoring target.
[0081] Step S302: multiple local tracks of a specific object are identified from the monitoring data processed by each positioning source server, where the specific object is any monitoring target.
[0082] It should be noted that the monitoring data processed by each positioning source server includes a local track of at least one monitoring target; the monitoring data processed by different positioning source servers may include a local track of the same monitoring target.
[0083] In the specific implementation of step S302 , a trajectory set is constructed using the monitoring data processed by each positioning source server. The trajectory set includes all local trajectories of all monitoring targets.
[0084] The constructed trajectory set is classified to obtain at least one classification cluster, each classification cluster contains all local trajectories of the same monitoring target; any classification cluster is obtained to obtain multiple local trajectories of a specific object, the specific object is any monitoring target, and the multiple local trajectories of the specific object are derived from monitoring data of different positioning source servers.
[0085] That is to say, the local trajectories of the monitoring targets in the monitoring data processed by all positioning source servers are associated, that is, the local trajectories processed by all positioning source servers are classified to obtain multiple local trajectories of a specific object, and the number of classifications is the number of monitoring targets. Among them, "multiple local trajectories of a specific object" are multiple local trajectories of any monitoring target.
[0086] It should be noted that the above-mentioned method of “classifying the constructed trajectory set” can be implemented by using a specified trajectory association algorithm.
[0087] Through the content of step S302, multiple local tracks of each monitoring target are identified one by one from the monitoring data processed by all positioning source servers.
[0088] Step S303: Fusing multiple local trajectories of the specific object to obtain a tracking trajectory of the specific object.
[0089] In the specific implementation of step S303, after identifying multiple local trajectories of a specific object (i.e., after identifying multiple local trajectories of each monitored target), the multiple local trajectories of the specific object are fused to obtain a tracking trajectory of the specific object with higher accuracy;
[0090] That is to say, the local trajectories of the same monitoring target in the monitoring data processed by all positioning source servers are fused to form a more accurate tracking trajectory representing each monitoring target.
[0091] In some embodiments, a specific method of fusing multiple local trajectories of a specific object is to perform weighted summation on the multiple local trajectories of the specific object using weight coefficients corresponding to each positioning source server to obtain a tracking trajectory of the specific object.
[0092] It should be noted that, since the multiple local trajectories of a specific object are derived from the monitoring data of each positioning source server, the multiple local trajectories of the specific object are weighted and summed by the weight coefficients corresponding to each positioning source server to obtain the tracking trajectory of the specific object.
[0093] It should be further explained that the accuracy of the local trajectories from each positioning source server is not equal (the measurement accuracy of each positioning source is different). In order to weigh the accuracy of each local trajectory of a specific object, this solution fuses multiple local trajectories of the specific object by marking the weight coefficient of the measurement accuracy (denoted as W);
[0094] Among them, the positioning source server with higher measurement accuracy has smaller error of local trajectory, and the corresponding weight coefficient of the positioning source server is larger;
[0095] On the contrary, the lower the measurement accuracy of the positioning source server, the greater the error of the local trajectory obtained, and the smaller the weight coefficient corresponding to the positioning source server.
[0096] like Figure 5 As shown in the example diagram of the weighted summation principle, the local trajectories of the specific object obtained by processing different positioning source servers are multiplied by the weight coefficients corresponding to the positioning source servers for weighted summation, thereby fusing the local trajectories of the specific object to obtain the tracking trajectory of the specific object; Figure 5 In, X 1 To X n are the local trajectories of specific objects processed by the 1st positioning source server to the nth positioning source server, W 1 To W n They are the weight coefficients corresponding to the 1st positioning source server to the nth positioning source server respectively.
[0097] Specifically, the local trajectory of a specific object processed by the i-th positioning source server is recorded as X i (i=1,2,…,n), X i The ith positioning source server processes the positioning data of the ith positioning source;
[0098] Since the measurement accuracy of each positioning source is different, assuming that X i Normal distribution ,u i is the expected value, σ i is the mean square error (representing the measurement accuracy of the positioning source);
[0099] Mean square error σ i The larger the value is, the greater the discreteness of the data obtained when the positioning source measures the same parameter, that is, the lower the measurement accuracy of the positioning source; conversely, the higher the measurement accuracy of the positioning source.
[0100] The weight coefficient corresponding to the i-th positioning source server (the positioning source server that processes the positioning data of the i-th positioning source) is W i (i=1,2,…,n), then the weighted sum of multiple local trajectories of a specific object can be performed through formula (1) to obtain the tracking trajectory of the specific object.
[0101] (1)
[0102] In formula (1), is the tracking trajectory of a specific object obtained by fusion, X i is the local trajectory of a specific object processed by the i-th positioning source server, W i is the weight coefficient corresponding to the i-th location source server.
[0103] Step S304: based on the monitoring data processed by each positioning source server, the extended trajectory of the specific object is estimated.
[0104] In the specific implementation of step S304, the monitoring data processed by each positioning source server is used to estimate the extended trajectory of the specific object.
[0105] It should be noted that, with regard to the execution order of step S302, step S303 and step S304, "step S302 to step S303" and "step S304" can be executed simultaneously, or steps S302 to step S304 can be executed sequentially in a front-to-back order. The execution order of steps S302 to step S304 is not limited herein.
[0106] Through the contents of the above steps S301 to S304, the tracking trajectory and extended trajectory of each monitored target can be obtained.
[0107] In some embodiments, the tracking trajectory and the extended trajectory of a specific object are uploaded in sequence as a reference trajectory of the specific object, that is, the tracking trajectory is uploaded first and then the extended trajectory;
[0108] That is to say, for each monitoring target, the tracking trajectory and the extended trajectory of the monitoring target are uploaded in sequence as the reference trajectory of the monitoring target.
[0109] In the embodiment of the present invention, monitoring data processed by each positioning source server using a designated coordinated universal time for time synchronization is obtained, and multiple local trajectories of a specific object are identified. The multiple local trajectories of the specific object are fused to obtain a tracking trajectory of the specific object, and an extended trajectory of the specific object is estimated. The tracking trajectory and the extended trajectory are processed to supplement the motion state of the delayed time process, so as to improve the time accuracy and position accuracy of the monitoring, thereby improving the monitoring effect.
[0110] For the above-mentioned embodiments of the present invention Figure 3 In order to better explain the content of "fusing multiple local trajectories of a specific object to obtain a tracking trajectory of the specific object" mentioned in step S303, Figure 6 Give an example.
[0111] like Figure 6 As shown in the schematic diagram of the principle of obtaining the tracking trajectory of a specific object by fusion, it can be seen that the trajectory (here equivalent to the track) of the aircraft in the airport is monitored through positioning devices 1 to positioning devices n (the positioning devices can be tracking sensors), and the motion state of the monitoring target (referred to as the target) and the airport terrain information are also collected at the same time;
[0112] Positioning device 1 to positioning device n respectively send the collected positioning data of the monitoring target to the corresponding positioning source server for processing;
[0113] For the "trajectory processing" part, each positioning source server uses GPS Coordinated Universal Time for time synchronization. The positioning source server tracks the monitored target based on the positioning data to obtain the local trajectory of the monitored target;
[0114] For the "trajectory fusion" part, the local trajectories of the monitored target are associated to identify multiple local trajectories of the same monitored target (i.e., multiple local trajectories of a specific object); the multiple local trajectories of the same monitored target are fused to obtain the tracking trajectory of the monitored target (i.e., the tracking trajectory of the specific object).
[0115] The above is an example of fusing local trajectories. For the specific implementation principle, please refer to the above embodiment of the present invention. Figure 3 , I will not go into details here.
[0116] For the above-mentioned embodiments of the present invention Figure 3 Step S304 involves estimating an extended trajectory of a specific object. In order to improve the consistency of the monitoring system when a high latency problem occurs, the reference trajectory of the monitored target will be supplemented by an extended trajectory (denoted as ET). Before explaining how to estimate the extended trajectory, the extended trajectory will be explained first.
[0117] Assume that the test object on the road (for example, a vehicle) performs two basic vehicle behaviors: “lane keeping” and “lane changing”. tgt The composition of the extended trajectory (ET) in is shown in formula (2).
[0118] (2)
[0119] In formula (2), is the target state, is the vertical coordinate, is the longitudinal velocity, is the longitudinal acceleration, is the lateral speed.
[0120] in, , and The specific content of is shown in formula (3).
[0121] (3)
[0122] In formula (3), set and is 0, J can be obtained by solving the acceleration record; J is the rate of change of longitudinal acceleration, is the longitudinal velocity, is the longitudinal acceleration, △T ct is the time difference from the current state to the target state, △T ce is the time difference from the current state to the motion estimation state.
[0123] See also Figure 7 , which shows a flow chart of estimating an extended trajectory of a specific object provided by an embodiment of the present invention, including the following steps:
[0124] Step S701: based on the monitoring data processed by each positioning source server, at least one candidate trajectory of a specific object is estimated.
[0125] In the specific implementation of step S701 , at least one candidate trajectory corresponding to the specific object is estimated using the monitoring data processed by each positioning source server.
[0126] Step S702: Evaluate the collision risk and logical cost of the candidate trajectory.
[0127] In the specific implementation of step S702 , the collision risk and logical cost of each candidate trajectory corresponding to the specific object are evaluated.
[0128] Specifically, the collision risk of candidate trajectories can be evaluated based on the distance between a specific object and obstacles in a dynamic environment. This scheme adopts a grid-based collision estimation method to improve the computational efficiency of collision risk.
[0129] like Figure 8 As shown in the example diagram of decomposing trajectory data and environmental data into a grid, taking the monitored target as a vehicle as an example, a curvilinear coordinate system is used to convert the trajectory data and environmental data into a rectangular form and further decompose them into a grid; the collision risk f is calculated using the following formula (4): c , f c That is, it is equal to the sum of the last matrices (0,1).
[0130] (4)
[0131] In formula (4), D tra To monitor the trajectory data of the target, D env A collection of other moving objects around the monitoring target;
[0132] It should be noted that, combined with Figure 8 The content shown is visible, D tra Occupancy is determined by analyzing the intersection points on the grid, and the grids on both sides of the intersection will be identified as occupied areas;
[0133] For D env , the position of surrounding objects will be estimated at each sampling interval based on the motion estimation model, with a value of 1 if the grid is occupied and a value of 0 if the grid is not occupied.
[0134] The above is an explanation of the collision risk of candidate trajectories.
[0135] For the logical cost of candidate trajectories, this scheme establishes a logical cost to evaluate the possibility of realizing candidate trajectories in physical space. Candidate trajectories with low logical costs are more likely to be selected as extended trajectories (that is, selected as ET); the logical cost can be calculated based on the historical consistency f h and the subsequent trajectory smoothness f s Get, historical consistency f h and the subsequent trajectory smoothness f s The definition of is given in formula (5) and formula (6).
[0136] (5)
[0137] (6)
[0138] In formula (5) and formula (6), T cur The current time when the data is received, T est To estimate the time, is the longitudinal coordinate ET, is the longitudinal coordinate TT, is the transverse coordinate ET, is the transverse coordinate TT, is the smoothness, △T ce is the time difference from the current state to the motion estimation state, △T ct is the time difference from the current state to the target state.
[0139] Step S703: Establish a search space containing all candidate trajectories.
[0140] In the specific implementation of step S703, based on the evaluation results of the collision risk and the logical cost, a search space for the extended trajectory is established, and the search space includes all candidate trajectories of the specific object; the definition of the candidate trajectories (also referred to as candidate vectors) in the search space is shown in formula (7).
[0141] (7)
[0142] In formula (7), p j is the jth candidate trajectory in the search space, d tgt (j) is the lateral target state, △T ct (j) is the time difference from the current state to the target state.
[0143] Step S704: Based on the collision risk and the logical cost, search for candidate trajectories that meet specific conditions from the search space to obtain an extended trajectory of the specific object.
[0144] In the specific implementation process of step S704, based on the collision risk and logical cost of all candidate trajectories of the specific object, a candidate trajectory that meets the specific condition is searched from the search space, so as to obtain an extended trajectory of the specific object. The extended trajectory of the specific object is “the candidate trajectory that meets the specific condition is searched in the search space”.
[0145] Specifically, through the objective function (also called fitness function, which calculates the maximum convergence) and specific conditions, the candidate trajectory that meets the specific conditions is searched from the search space, so as to obtain the extended trajectory of the specific object;
[0146] The specific content of the objective function is shown in formula (8), and the specific content of the specific conditions is shown in formula (9).
[0147] (8)
[0148] (9)
[0149] In formula (8) and formula (9), fh and f s are historical consistency and subsequent trajectory smoothness, μ h and μ s The two smoothnesses are f h and f s The reference value, ω h and ω s are the weights of each preset indicator, f c is the collision risk, and j is the jth candidate trajectory in the search space.
[0150] Search the search space for candidate trajectories that satisfy formula (9), and the candidate trajectories that satisfy formula (9) are the extended trajectories of the specific object.
[0151] Through the contents of the above steps S701 to S704, the extended trajectory of each monitored target can be estimated.
[0152] The above embodiments of the present invention Figure 7 These are instructions on how to estimate the extended trajectory of a particular object.
[0153] In summary, this solution combines "real-time fusion of multi-source perception data" and "synchronous virtual and real simulation" to construct a multi-source trajectory fusion and virtual and real space synchronization processing process; for the aforementioned process, Fig. 9 As shown in the overall process example diagram of a method for processing monitoring data from multiple sources, the aforementioned process is divided into the following five core function modules: send Module, G(ξ est+fus ) module, G(TT) module, G(ET) module, theme update module.
[0154] like Fig. 9 As shown in the figure, the overall process includes five levels: receiving end, parsing, generating, planning and executing. Specifically, due to different network conditions, there are different network delays △T at the receiving end. The parsing end implements the parsing of data and executes the sending action.
[0155] Execute G(ξ est+fus ) module, and then executes the tasks of calculating TT (tracking trajectory) and ET (extended trajectory) in parallel. Among them, TT can be quickly solved by the G(TT) module, and the time consumption for solving TT is generally less than 1ms. The G(ET) module calculates candidate trajectories and selects the most suitable candidate trajectory as ET;
[0156] The planning layer is equivalent to a data container for storing the reference trajectory obtained from the generation layer. Fig. 9 It can be seen that TT is uploaded first, and then ET is uploaded;
[0157] The mobile object that executes the control mirror updates the subject according to the object from the planning layer, where "the object from the planning layer" is the extended trajectory and "updating the subject" means clearing the last cached trajectory.
[0158] In general, on the one hand, this scheme unifies the local trajectories measured by various positioning sources (such as synchronous sensors or asynchronous sensors) at different times to the same time on the time axis, and then associates the local trajectories to determine the local trajectory of the same monitoring target measured by different positioning sources, and then fuses the local trajectories of the same monitoring target to obtain a more accurate tracking trajectory;
[0159] On the other hand, in order to improve the continuity and accuracy of the virtual-real state under different communication delay interference and interaction delay interference, this scheme constructs a delay time model based on the interaction time delay of the communication and monitoring systems, and divides the virtual-real motion trajectory synchronization problem into two stages: tracking trajectory and extending trajectory;
[0160] Through the above two aspects, the time accuracy and location accuracy of monitoring can be effectively improved, thereby improving the monitoring effect.
[0161] Corresponding to the method for processing monitoring data from multiple sources provided by the above embodiment of the present invention, see Fig.10 , an embodiment of the present invention also provides a structural block diagram of a device for processing monitoring data from multiple sources, the device comprising: an acquisition unit 100, an identification unit 200, a fusion unit 300 and an estimation unit 400;
[0162] The acquisition unit 100 is used to acquire monitoring data processed by each positioning source server. Each positioning source server uses a designated coordinated universal time for time synchronization. The monitoring data at least includes a local trajectory of the monitoring target.
[0163] The identification unit 200 is used to identify multiple local tracks of a specific object from the monitoring data processed by each positioning source server, and the specific object is any monitoring target.
[0164] The fusion unit 300 is used to fuse multiple local trajectories of a specific object to obtain a tracking trajectory of the specific object.
[0165] In a specific implementation, the fusion unit 300 is specifically used to: use the weight coefficients corresponding to each positioning source server to perform weighted summation on multiple local trajectories of a specific object to obtain the tracking trajectory of the specific object.
[0166] The estimation unit 400 is used to estimate the extended trajectory of a specific object based on the monitoring data processed by each positioning source server.
[0167] In the embodiment of the present invention, monitoring data processed by each positioning source server using a designated coordinated universal time for time synchronization is obtained, and multiple local trajectories of a specific object are identified. The multiple local trajectories of the specific object are fused to obtain a tracking trajectory of the specific object, and an extended trajectory of the specific object is estimated. The tracking trajectory and the extended trajectory are processed to supplement the motion state of the delayed time process, so as to improve the time accuracy and position accuracy of the monitoring, thereby improving the monitoring effect.
[0168] Preferably, combined Fig.10 As shown in the figure, the identification unit 200 includes a construction subunit, a classification subunit and an acquisition subunit, and the execution principle of each subunit is as follows:
[0169] The construction subunit is used to construct a trajectory set using the monitoring data processed by each positioning source server.
[0170] The classification subunit is used to classify the trajectory set to obtain at least one classification cluster, each classification cluster containing all local trajectories of the same monitoring target.
[0171] The acquisition subunit is used to obtain any classification cluster to obtain multiple local trajectories of a specific object.
[0172] Preferably, combined Fig.10 As shown in the figure, the estimation unit 400 includes an estimation subunit, an evaluation subunit, a creation subunit and a search subunit, and the execution principle of each subunit is as follows:
[0173] The estimation subunit is used to estimate at least one candidate trajectory of a specific object based on the monitoring data processed by each positioning source server.
[0174] The evaluation subunit is used to evaluate the collision risk and logical cost of candidate trajectories.
[0175] A subunit is established to establish a search space containing all candidate trajectories.
[0176] The search subunit is used to search for candidate trajectories that meet specific conditions from the search space based on collision risk and logical cost to obtain an extended trajectory of a specific object.
[0177] Preferably, combined Fig.10 The device also includes:
[0178] The uploading unit is used to sequentially upload the tracking trajectory and the extended trajectory as a reference trajectory of a specific object.
[0179] In summary, the embodiment of the present invention provides a method and device for processing monitoring data from multiple sources, obtains monitoring data processed by each positioning source server that uses a specified coordinated universal time for time synchronization, and identifies multiple local trajectories of a specific object. The multiple local trajectories of the specific object are fused to obtain the tracking trajectory of the specific object, and the extended trajectory of the specific object is estimated. The tracking trajectory and the extended trajectory are processed to supplement the motion state of the delayed time process, so as to improve the time accuracy and position accuracy of the monitoring, thereby improving the monitoring effect.
[0180] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.
[0181] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0182] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for processing monitoring data from multiple sources, characterized in that: The method comprises: Acquire monitoring data processed by each positioning source server, each positioning source server adopts a designated coordinated universal time for time synchronization, and the monitoring data at least includes a local trajectory of the monitoring target; From the monitoring data processed by each of the positioning source servers, a plurality of local trajectories of a specific object are identified, the specific object being any of the monitoring targets; Fusion of the multiple local trajectories of the specific object to obtain the tracking trajectory of the specific object includes: using weight coefficients corresponding to the respective positioning source servers to perform weighted summation on the multiple local trajectories of the specific object to obtain the tracking trajectory of the specific object; Based on the monitoring data processed by each of the positioning source servers, estimating the extended trajectory of the specific object, including: based on the monitoring data processed by each of the positioning source servers, estimating at least one candidate trajectory of the specific object; evaluating the collision risk and logical cost of the candidate trajectory; establishing a search space containing all the candidate trajectories; based on the collision risk and the logical cost, searching the search space for the candidate trajectory that meets specific conditions to obtain the extended trajectory of the specific object; the logical cost is based on historical consistency and subsequent trajectory smoothness obtained, , ,in, is the current time when the data is received, To estimate the time, is the target state time, is the ET longitudinal coordinate, is the longitudinal coordinate of TT, is the ET transverse coordinate, is the TT horizontal coordinate, is the smoothness, is the time difference from the current state to the motion estimation state, is the time difference from the current state to the target state, TT is the time difference from ξ cur To est+fus The reference trajectory is called ET from ξ est+fus To tgt The reference trajectory is called ξ cur is the virtual space data receiving state, ξ est+fus is the motion estimation state, ξ tgt is the target state; The tracking trajectory and the extended trajectory are uploaded in sequence as a reference trajectory of the specific object.
2. The method according to claim 1, characterized in that From the monitoring data processed by each of the positioning source servers, a plurality of local trajectories of a specific object are identified, including: Using the monitoring data processed by each of the positioning source servers, construct a trajectory set; Classifying the trajectory set to obtain at least one classification cluster, each of which contains all local trajectories of the same monitoring target; Any of the classification clusters is obtained to obtain a plurality of local trajectories of a specific object.
3. A device for processing monitoring data from multiple sources, characterized in that: The device comprises: An acquisition unit, configured to acquire monitoring data processed by each positioning source server, each positioning source server adopting a designated coordinated universal time for time synchronization, and the monitoring data at least including a local track of the monitoring target; an identification unit, configured to identify a plurality of local tracks of a specific object from the monitoring data processed by each of the positioning source servers, wherein the specific object is any monitoring target; A fusion unit, used for fusing multiple local trajectories of the specific object to obtain a tracking trajectory of the specific object; The fusion unit is specifically used to: perform weighted summation on multiple local trajectories of the specific object using weight coefficients corresponding to each of the positioning source servers to obtain a tracking trajectory of the specific object; an estimating unit, configured to estimate an extended trajectory of the specific object based on the monitoring data processed by each of the positioning source servers; The estimation unit includes: an estimation subunit, an evaluation subunit, a creation subunit and a search subunit; The estimation subunit is configured to estimate at least one candidate trajectory of the specific object based on the monitoring data processed by each of the positioning source servers; The evaluation subunit is used to evaluate the collision risk and logical cost of the candidate trajectory; the logical cost is based on historical consistency and subsequent trajectory smoothness obtained, , ,in, is the current time when the data is received, To estimate the time, is the target state time, is the ET longitudinal coordinate, is the longitudinal coordinate of TT, is the ET transverse coordinate, is the TT horizontal coordinate, is the smoothness, is the time difference from the current state to the motion estimation state, is the time difference from the current state to the target state, TT is the time difference from ξ cur To est+fus The reference trajectory is called ET from ξ est+fus To tgt The reference trajectory is called ξ cur is the virtual space data receiving state, ξ est+fus is the motion estimation state, ξ tgt is the target state; The establishing subunit is used to establish a search space containing all the candidate trajectories; The search subunit is configured to search the candidate trajectory satisfying a specific condition from the search space based on the collision risk and the logical cost, so as to obtain an extended trajectory of the specific object; The uploading unit is used to sequentially upload the tracking trajectory and the extended trajectory as a reference trajectory of the specific object.
4. The device according to claim 3, characterized in that The identification unit comprises: A construction subunit, configured to construct a trajectory set using the monitoring data processed by each of the positioning source servers; A classification subunit, used for classifying the trajectory set to obtain at least one classification cluster, each of which contains all local trajectories of the same monitoring target; The acquisition subunit is used to acquire any of the classification clusters to obtain multiple local trajectories of a specific object.
