CN-121660401-B - Road heavy goods vehicle NO based on improved hidden Markov modelXEmission calculation method

CN121660401BCN 121660401 BCN121660401 BCN 121660401BCN-121660401-B

Abstract

The invention relates to the field of urban road traffic emission assessment, in particular to a method for calculating NO X emission of a road heavy truck based on an improved hidden Markov model, which comprises the following steps of S1, identifying internal and external tracks of a road and extracting tracks on the road, S2, segmenting tracks based on datum line offset, S3, extracting track key points based on offset distance statistics, S4, reducing a candidate road network based on micro-area segmentation, S5, matching tracks based on the hidden Markov model, S6, calculating the running condition of a road vehicle, and S7, calculating and summarizing the NO X emission of the road vehicle. The method optimizes the calculation flow of the traditional hidden Markov model, and obviously reduces the calculation complexity while ensuring the track matching precision. Compared with the prior art, the method and the device can effectively solve the problem of low road-level emission calculation efficiency under the condition of low sampling frequency GPS data, and enable the road traffic pollution emission simulation process to be more efficient.

Inventors

LIU HAOBING
GAO PENGFEI

Assignees

同济大学

Dates

Publication Date: 20260505
Application Date: 20260205

Claims (7)

1. The method for calculating the NO X emission of the road heavy truck based on the improved hidden Markov model is characterized by comprising the following steps of: S1, extracting effective on-track through road inside and outside track identification based on GPS track data of original heavy diesel trucks ; S2, segmenting the effective track based on a datum line offset criterion to enable the heading of the continuous GPS points in the sub-track to be consistent with the datum line direction; S3, extracting key motion nodes aiming at all sub-tracks to realize self-adaptive simplification of GPS track data; S4, reducing candidate road networks based on a micro-region splicing method, and constructing a simplified road network for a hidden Markov model; s5, matching the simplified track and the simplified road network based on the hidden Markov model, and reconstructing a vehicle running path; S6, calculating the operation condition of the vehicle at the road level by using the track matching result; S7, calculating the single vehicle road grade emission by combining the vehicle attribute and the road grade operation condition, and integrating to form the urban road grade NO X emission distribution; The step S4 is specifically performed by, Constructing a micro-area corresponding to each sub-track section obtained in the step S2 for limiting a candidate road range for vehicle running, wherein the width of the micro-area is set along the direction perpendicular to the sub-track, the size of the micro-area is determined by the maximum transverse displacement between adjacent track points, and the transverse displacement is constrained by the maximum running speed of the vehicle, the GPS sampling time interval and a preset expansion parameter epsilon, namely Wherein, the Representing a micro-region width corresponding to the ith sub-track; Representing the maximum running speed between adjacent track points in the ith sub-track, delta t representing the sampling time interval of GPS track data, epsilon representing the expansion margin introduced for ensuring that a feasible path exists between the adjacent track points; the length of the micro-region is set along the direction of the sub-track, and the size of the micro-region is determined by the distance between the starting point and the end point of the sub-track segment and combined with an expansion parameter epsilon, namely: Wherein, the Representing a micro-region length corresponding to the ith sub-track; Representing the direction along the sub-track, from the start point of the sub-track And end point A distance between the two endpoints; On the basis, the direction constraint is applied to the initial road network, and only the included angle between the road section direction and the sub-track section direction is kept not larger than the preset maximum allowable angle threshold Combining the candidate road sets corresponding to the sub-track segments to obtain a reduced candidate road network The method comprises the following steps: Wherein, the Representing an initial road network, e representing a reduced road network Road segments of (a); representing micro-regions corresponding to the ith sub-track, N representing the total number of sub-tracks; An angular deviation between the direction of the section e and the direction of the sub-track reference line l m,n is represented; representing a maximum allowable angular deviation threshold; The step S6 is specifically performed by, The effective track sequence obtained in the step S1 is adopted And the optimal matching path S * obtained in the step S5, a projection sequence corresponding to the track sequence on the matching path is obtained, namely: Wherein, the Representing the sequence of trajectories after projection, proj (·) representing orthogonal projection of the trajectory points to the matching road path A projection operator on the image; for any road section Defining a corresponding projection point index set: Wherein, the Representing road segments Is provided with a projection point index of (c), Representing the projected trajectory point, The representation being located in a road section Index set of projection points on the index set, and the base number is recorded as M is the optimal matching path sequence length; On road section When the number of the projection points on the road section is less than two, estimating the operation condition of the road section based on the boundary projection points of the adjacent road section; For road sections with less than two projection points, selecting start-stop boundary points for calculating operation conditions: Wherein, the Representing the last proxel of the immediately preceding adjacent road segment, A first proxel representing a subsequent adjacent road segment; For road sections with the number of projection points not less than two, identifying an idle track subsequence according to the space distance between adjacent projection points, wherein when the distance between adjacent projection points is smaller than a preset distance threshold value, the corresponding track section is judged to be in an idle state, otherwise, the corresponding track section is judged to be in a normal driving state, and the track sequence is identified as follows: Wherein, the And Respectively representing an idle subsequence and a driving subsequence which are obtained by recognition on a road section; representing the spatial distance between adjacent projected trajectory points; is a preset distance threshold; Calculating a segmented average travel speed sequence for a road segment : Wherein D (-) and Δt (-) respectively represent the driving distance along the matched road track and the corresponding time interval; indicating an idle speed value.
2. The method for calculating the emissions of a road heavy goods vehicle NO X based on the improved hidden markov model according to claim 1, wherein step S1 is specifically, An original GPS track sequence O= { O 1 ,o 2 ,…,o T } of the urban heavy diesel truck is collected, and for each GPS track point in the track sequence, the GPS track point and a road network are calculated The minimum space distance between the two track points is determined to be the track point when the minimum distance is smaller than a preset distance threshold gamma, otherwise, the track point is determined to be the off-track point, namely the track sequence in the track should satisfy the following conditions: Wherein, the Representing GPS track point o i and road network The minimum spatial distance between the two, gamma being a distance threshold for distinguishing between the track points and the track points, Is the length of the track point sequence; removing abnormal position points caused by GPS positioning errors, thereby obtaining effective track sequence 。
3. The improved hidden markov model based road heavy goods vehicle NO X emission calculating method according to claim 2, wherein when the time interval between the occurrence of a track point in a track and its adjacent track point is not more than a preset time threshold value, it is regarded as a position outlier caused by a GPS positioning error.
4. The method for calculating the emissions of a road heavy goods vehicle NO X based on the improved hidden markov model according to claim 1, wherein step S2 is specifically, The effective track sequence obtained in step S1 For input, starting from the third track point in the track sequence, traversing and identifying track segment points, wherein the traversing flow is as follows: Starting with the current sub-track sequence Endpoint with The connecting line is used as a datum line When the offset distance between any track point and the datum line exceeds the preset threshold delta, the track point is judged to be a new segment point and updated to be a new sub track starting point to continue traversing, and the track starting point, the end point and the segment point identified in the traversing process jointly form a sub track boundary point set Namely, the boundary points should satisfy: wherein delta is a preset distance threshold; And (3) with Respectively representing a starting point and an ending point of the sub-track; Representing the starting point And end point A reference line is formed; Representing the locus o i and the datum line A reference line offset distance between the track points o i and the reference line Is a vertical distance of (c).
5. The method for calculating the emissions of a road heavy goods vehicle NO X based on the improved hidden markov model according to claim 1, wherein step S3 is specifically, For each sub-track segment obtained in step S2, using the starting point of the sub-track segment And end point For the middle track point in the sub track section, calculating the offset angle of the middle track point relative to the reference line, and combining the geographic distance between adjacent track points to obtain the offset distance of the track point, wherein the calculation formula is as follows: Wherein, the Representing a trace point Relative to a reference line Is a distance of offset; Representing a geographic distance between adjacent GPS observation points; Representing a datum line Direction and trajectory point of (2) To the point of Angular deviation between the directions of motion of (a); Calculating the average value of the offset distances of all the track points of each sub-track segment, identifying the track point as a key movement point when the offset distance of a certain track point is larger than the average offset distance of the corresponding sub-track segment, and reserving the starting point, the end point and the identified key movement point of each sub-track segment to form a simplified track sequence Each trace point satisfies: Wherein, the Represents the average of the offset distances of all the track points within the sub-track defined by the reference line.
6. The method for calculating the emissions of a road heavy goods vehicle NO X based on the improved hidden markov model according to claim 1, wherein step S5 is specifically, Simplified track sequence based on step S3 And the candidate road network obtained in step S4 Constructing a hidden Markov model including an initial state probability vector State transition probability matrix And an observation probability matrix Solving the hidden Markov model by adopting a Viterbi algorithm to obtain a hidden state sequence with maximum posterior probability, wherein the hidden state sequence corresponds to an optimal matching path of a vehicle on a road network The method comprises the following steps: 。
7. The method for calculating the emissions of a road heavy goods vehicle NO X based on the improved hidden Markov model according to claim 1, wherein the step S7 is specifically, The segmented average running speed sequence obtained in the step S6 is adopted And calculating an emission of the road segment in combination with the vehicle attribute information and the speed-related emission factor corresponding to the emission standard to perform calculation: When (1): When (1): Wherein, the Represents the amount of emissions from the road segment, Indicating a speed dependent emission factor at emission standard x when the continuous idle duration exceeds a preset time threshold When the engine is judged to be in a flameout state, the corresponding emission amount is set to be zero; Accumulating the corresponding discharge amounts of all vehicle tracks passing through the same road section in the same time interval to obtain the total discharge amount of the road section in the time interval; and summarizing the emission of all road sections in the urban area range to form an urban area scale road level vehicle emission distribution result.

Description

Road heavy truck NO X emission calculation method based on improved hidden Markov model Technical Field The invention relates to the field of urban road traffic emission assessment, in particular to a road heavy goods vehicle NO X emission calculation method based on an improved hidden Markov model. Background With the acceleration of the urban process and the continuous increase of the logistics transportation demand, the duty ratio of the heavy diesel trucks in the urban road traffic system is continuously increased. Because the heavy diesel trucks have the characteristics of large vehicle mass, high engine power and large emission intensity, the emitted nitrogen oxides become important components of urban traffic source air pollution, and have obvious influence on urban air quality improvement and public health. Therefore, the refined evaluation of the NO X emission of the commercial heavy diesel trucks is developed, and the method has important significance for traffic pollution prevention, control and management decision. The existing road traffic emission assessment method mainly comprises a macroscopic emission inventory method based on statistical data and a microscopic emission estimation method based on vehicle track data. The macroscopic emission list method generally depends on traffic investigation data, road grades and experience emission factors, estimates emission in areas or road grade layers, and is difficult to reflect differences of vehicle operation conditions among different road sections, and spatial resolution is low. In contrast, the emission assessment method based on GPS track data can characterize the real running state of the vehicle, and provides a data base for road-level emission assessment of even finer scales. In the road-level emission evaluation process based on GPS track data, the rapid and accurate matching of the track and the road network is a key premise. At present, the hidden Markov model is widely applied to map matching of vehicle tracks because the observation error and the road topology structure can be comprehensively considered. However, when the conventional HMM map matching method processes large-scale vehicle track data in the market domain scale, the problems of high computational complexity, large candidate road network scale, low matching efficiency and the like are often faced, and the situations of unstable matching or insufficient computational efficiency are likely to occur. In addition, aiming at the urban road grade emission distribution calculation of large-scale GPS track data, the existing method still has the defect in terms of calculation efficiency, and the actual requirements of refined traffic emission assessment and dynamic management are difficult to meet. Therefore, a method for improving the track road matching efficiency and stability and realizing the rapid and refined calculation of the urban road grade NO X emission under the condition of GPS track data of a low sampling frequency heavy diesel truck is needed to support the urban traffic pollution emission assessment and management decision. Disclosure of Invention The invention aims to solve the technical problem of providing an improved hidden Markov model-based method for calculating the NO X emission of a road heavy diesel truck, which can quickly and accurately generate the NO X emission distribution of the urban road heavy diesel truck based on GPS track data flow. Technical proposal The method for calculating the NO X emission of the road heavy truck based on the improved hidden Markov model comprises the following steps of: s1, extracting effective on-road tracks through the identification of the inner and outer tracks of a road based on the GPS track data of an original heavy diesel truck; s2, segmenting the effective track based on a datum line offset criterion to enable the heading of continuous GPS points in the sub-track to be consistent with the datum line; S3, extracting key motion nodes aiming at all sub-tracks to realize self-adaptive simplification of GPS track data; S4, reducing candidate road networks based on a micro-region splicing method, and constructing a simplified road network for a hidden Markov model; s5, matching the simplified track and the simplified road network based on the hidden Markov model, and reconstructing a vehicle running path; S6, calculating the operation condition of the vehicle at the road level by using the track matching result; And S7, calculating the single vehicle road grade emission by combining the vehicle attribute and the road grade operation condition, and integrating to form the urban road grade NO X emission distribution. In step S1, an original GPS track sequence o= { O 1,o2,…,oT }, for each GPS track point in the track sequence, is collected, and calculated from the original GPS track sequence and the road networkA minimum spatial distance therebetween. And when the minimum distance is smaller than a preset distance thres