CN-122024510-A - Multi-source speed limit information fusion decision method and system based on rules

CN122024510ACN 122024510 ACN122024510 ACN 122024510ACN-122024510-A

Abstract

The invention provides a rule-based multi-source speed limit information fusion decision method and system, which comprises the steps of S1, obtaining multi-source speed limit information comprising visual speed limit information, navigation speed limit information and vehicle state information, S2, carrying out validity judgment on the visual speed limit information based on a preset validity judgment rule, S3, constructing a candidate speed limit set by using the valid visual speed limit information and the navigation speed limit information based on a validity judgment result, S4, taking the real speed limit of a current road as an implicit state variable, taking the valid visual speed limit information and the navigation speed limit information as observation information of the implicit state variable, and constructing a speed limit state inference model comprising time sequence constraint, S5, carrying out confidence assessment on each candidate speed limit state at the current moment and obtaining the result based on the model under the constraint of the candidate speed limit set by combining a speed limit state assessment result at the previous moment and time sequence constraint, and S6, outputting a final speed limit value under the current road condition.

Inventors

ZHOU JINSUO
KONG DEYANG
LUO CHAGEN

Assignees

同济大学

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (9)

1. A multi-source speed limit information fusion decision method based on rules is characterized by comprising the following steps: s1, multi-source speed limit information acquisition, namely acquiring multi-source speed limit information, wherein the multi-source speed limit information comprises visual speed limit information, navigation speed limit information and vehicle state information; S2, a visual input validity judging step, namely, carrying out validity judgment on the visual speed limit information based on a preset validity judging rule to obtain a validity judging result; S3, constructing a candidate speed limit set by using the effective visual speed limit information and the navigation speed limit information based on the effectiveness judgment result; s4, constructing a speed limit state inference model, namely constructing a speed limit state inference model containing time sequence constraint by taking the real speed limit of the current road as an implicit state variable and taking the effective visual speed limit information and the navigation speed limit information as observation information of the implicit state variable; s5, carrying out confidence assessment on each candidate speed limit state at the current moment by combining the speed limit state assessment result at the previous moment with the time sequence constraint under the constraint of the candidate speed limit set based on the speed limit state inference model to obtain a confidence result of each candidate speed limit state; And S6, outputting, namely determining and outputting a final speed limit value under the current road condition according to the confidence coefficient result.
2. The rule-based multi-source speed limit information fusion decision method of claim 1, wherein: Wherein in the step S1, the visual speed limit information is road traffic speed limit identification information, The vehicle state information includes position information, travel speed, travel mileage, and travel time of the vehicle.
3. The rule-based multi-source speed limit information fusion decision method of claim 2, wherein: in S2, the method for obtaining the validity judgment result includes: the validity judgment rules comprise general condition speed limit scene rules: when the speed limit value in the visual speed limit information is larger than the current road speed limit value contained in the navigation speed limit information, judging that the visual speed limit information is invalid; When the speed limit value in the visual speed limit information is equal to the current road speed limit value contained in the navigation speed limit information or is equal to the default speed limit value determined according to the road type in the navigation speed limit information, judging that the visual speed limit information is valid; In other cases, it is determined that the visual rate-limiting information is invalid.
4. The rule-based multi-source speed limit information fusion decision method of claim 2, wherein: in S2, the method for obtaining the validity judgment result includes: The validity judgment rules comprise regional condition speed limit scene rules: when the regional condition speed limit value in the visual speed limit information is consistent with the regional condition speed limit value contained in the navigation speed limit information, setting a first effective distance threshold and a first effective time threshold; When the regional condition speed limit value in the visual speed limit information is inconsistent with the regional condition speed limit value contained in the navigation speed limit information, setting a second effective distance threshold and a second effective time threshold, wherein the second effective distance threshold and the second effective time threshold are smaller than the first effective distance threshold and the first effective time threshold; calculating an accumulated travel distance and an accumulated travel time of the vehicle from road traffic speed limit identification information of the identification area condition based on the travel mileage and the travel time in the vehicle state information; And when the accumulated running distance exceeds the effective distance threshold corresponding to the consistent or inconsistent situation or the accumulated running time exceeds the effective time threshold corresponding to the consistent or inconsistent situation, judging that the regional condition speed limit information is invalid, and otherwise, judging that the regional condition speed limit information is valid.
5. The rule-based multi-source speed limit information fusion decision method of claim 1, wherein: in the step S3, the method for constructing the candidate speed limit set includes: Constructing a candidate speed limit set based on effective visual speed limit information, navigation speed limit information, road type information contained in the navigation speed limit information and speed limit scene information, wherein the candidate speed limit set is used for representing a speed limit value range possibly established under the current road condition; And screening the speed limit information in the candidate speed limit set, and eliminating the speed limit information which does not meet the current road condition or traffic regulation requirement.
6. The rule-based multi-source speed limit information fusion decision method of claim 1, wherein: In S5, the time sequence constraint is implemented by a condition that the speed limit state change between adjacent time points meets a condition that the speed limit state change amplitude does not exceed a preset threshold value, so as to inhibit speed limit fluctuation caused by single abnormal speed limit observation.
7. The rule-based multi-source speed limit information fusion decision method of claim 6, wherein: in S5, when evaluating the confidence level of each candidate speed limit state at the current moment, the candidate speed limit state with the state change amplitude exceeding the preset threshold value is given a lower confidence level.
8. The rule-based multi-source speed limit information fusion decision method of claim 1, wherein: In S6, when the confidence coefficient results of the candidate speed limit states are all lower than the preset confidence coefficient threshold value, the speed limit value at the previous moment is output as the final speed limit value under the current road condition.
9. A rule-based multi-source speed limit information fusion decision system, comprising: The system comprises a multi-source speed limit information acquisition module, a speed limit information acquisition module and a speed limit information processing module, wherein the multi-source speed limit information comprises visual speed limit information, navigation speed limit information and vehicle state information; The visual input validity judging module is used for carrying out validity judgment on the visual speed limit information based on a preset validity judging rule to obtain a validity judging result; the candidate speed limit set construction module is used for constructing a candidate speed limit set by using the effective visual speed limit information and the navigation speed limit information based on the effectiveness judgment result; The speed limit state inference model construction module is used for constructing a speed limit state inference model containing time sequence constraint by taking the real speed limit of the current road as an implicit state variable and taking the effective visual speed limit information and the navigation speed limit information as observation information of the implicit state variable; The confidence evaluation module is used for carrying out confidence evaluation on each candidate speed-limiting state at the current moment by combining the speed-limiting state evaluation result at the previous moment with the time sequence constraint under the constraint of the candidate speed-limiting set based on the speed-limiting state inference model to obtain a confidence result of each candidate speed-limiting state; And the output module determines and outputs a final speed limit value under the current road condition according to the confidence coefficient result.

Description

Multi-source speed limit information fusion decision method and system based on rules Technical Field The invention relates to the technical field of intelligent driving and Advanced Driving Assistance Systems (ADAS) of vehicles, in particular to a multi-source speed limit information fusion decision method and system based on rules. Background According to the related traffic safety statistics, overspeed driving is a key factor of deadly accidents exceeding 25% per year, so that intelligent speed assisting systems for assisting drivers to maintain proper speeds are applied, such as a speed limiting system based on visual perception, a correct speed limiting value of a current road environment is determined by identifying speed limiting plate information on a road, or a speed limiting system based on vehicle navigation, road information is acquired in advance through a map merchant, and speed limiting value information of the current road is informed to the drivers through a vehicle machine system. Because of more traffic marks, different areas have different, partial roads have no traffic marks, the navigation system is not updated timely, urban and street scenes are positioned inaccurately, and the vehicle system is accidentally faulty, the requirements of accuracy and stability of the intelligent speed limiting system in practical application can not be met by only relying on a single vision system or the navigation system. Disclosure of Invention The invention aims to solve the problems and aims to provide a multi-source speed limit information fusion decision method and system based on rules. The invention provides a rule-based multi-source speed limit information fusion decision method which is characterized by comprising the following steps of S1, a multi-source speed limit information acquisition step, S2, a visual input validity judgment step, a candidate speed limit set construction step, S4, a speed limit state inference model construction step, a confidence evaluation step and a step 6, wherein the multi-source speed limit information acquisition step is used for acquiring multi-source speed limit information, the multi-source speed limit information comprises visual speed limit information, navigation speed limit information and self-vehicle state information, the visual input validity judgment step is used for carrying out validity judgment on the visual speed limit information based on a preset validity judgment rule to acquire validity judgment results, the candidate speed limit set construction step is used for constructing a candidate speed limit set by using the valid visual speed limit information and the navigation speed limit information based on the validity judgment results, the speed limit state inference model is constructed by taking the valid visual speed limit information and the navigation speed limit information as observation information of the hidden state variables, the confidence evaluation step is used for constructing a speed limit state inference model comprising time sequence constraint, the confidence evaluation step 5 is used for carrying out confidence evaluation on the basis of the speed limit state inference model, the confidence evaluation step is used for carrying out confidence evaluation on each confidence state at each time under the constraint of the candidate speed limit set, the confidence evaluation result is carried out confidence evaluation on each candidate speed limit at each time, the current time is obtained, and the final road condition is determined, and the current road condition is output. The rule-based multi-source speed limit information fusion decision method provided by the invention can be characterized in that the visual speed limit information is road traffic speed limit identification information, and the vehicle state information comprises the position information, the running speed, the running mileage and the running time of the vehicle. The multi-source speed limit information fusion decision method based on the rules is characterized in that the availability judgment result is obtained by the method that the availability judgment rule comprises a general condition speed limit scene rule, the visual speed limit information is judged to be invalid when the speed limit value in the visual speed limit information is larger than the current road speed limit value contained in the navigation speed limit information, the visual speed limit information is judged to be valid when the speed limit value in the visual speed limit information is equal to the current road speed limit value contained in the navigation speed limit information or equal to the default speed limit value determined according to the road type in the navigation speed limit information, and the visual speed limit information is judged to be invalid otherwise. The multi-source speed limit information fusion decision method based on the rules is characterized in that in S2