Intelligent sensing control system for riding protection equipment

Abstract

The invention discloses an intelligent sensing control system for riding protection equipment, relates to the technical field of riding safety protection, and is used for solving the problem that inertial sideslip is inaccurately intervened in advance. The method is composed of riding data collection, behavior event recognition, behavior matching, risk evolution analysis and pre-response control, the system constructs a behavior map through a multi-source sensor, the behavior map and an inertia sideslip template are subjected to feature difference value and structure consistency verification, and the behavior map is obtained by combining a trend change mode and a minimum continuous threshold value. Pre-response (pre-inflation, prompt rhythm and optical signal adjustment) is triggered at the early risk stage, closed-loop comparison is performed on the execution state and the risk state, and compared with a single-threshold triggering scheme, false alarm and missing alarm can be reduced, the response lead and intervention accuracy can be improved, unnecessary triggering can be reduced, and the self-adaptive reliability under a complex road can be enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of riding safety protection, and more particularly to an intelligent perception control system for riding protection equipment. BACKGROUND

[0002] With the popularity of motorcycles, electric bicycles and shared riding, the proportion of riding in urban and rural transportation systems is increasing year by year, but the traffic safety problems brought by it are also increasingly prominent. The rider lacks protection from the external structure of the vehicle in an open road environment, and once imbalance, side slip or collision occurs, it is easy to cause injury, especially in the case of rain, snow, slippery road surface, sharp curve section or emergency avoidance, the probability of high-risk behaviors such as inertial side slip and overturning significantly increases, and the accident develops quickly, and the controllable time is very short. The conventional method of relying on the rider's own reaction to avoid is difficult to effectively avoid injury.

[0003] The existing technology has the following problems: the existing riding protection equipment relies on a single trigger condition to start the protection action, for example, a collision sensor, an inclination sensor or an acceleration threshold value is used to trigger an airbag or a prompt system. This method often needs to react when the risk event is close to or has entered an uncontrollable stage, resulting in a delayed response and difficulty in reducing accident risks in a timely manner. In addition, although some systems introduce multiple sensors, they lack multi-dimensional data fusion analysis and trend prediction capabilities for complex riding behaviors, and cannot early identify high-risk behaviors such as inertial side slip that evolve gradually. There is a high rate of missed and false judgments, and it is impossible to achieve precise pre-response before the risk is fully revealed. There is a lack of closed-loop verification and correction mechanism between the protection strategy and the execution action, resulting in a dependence on a single judgment result for the protection effect, and insufficient stability and reliability. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the following scheme is provided to solve the problem of inaccurate inertial side slip early intervention in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: An intelligent perception control system for riding protection equipment, comprising a riding data acquisition module, a behavior event identification module, a behavior matching module, a risk evolution analysis module and a pre-response control module, and the modules are connected through signals; The riding data acquisition module is used to acquire time sequence signals containing acceleration, angular velocity, speed and trajectory points according to the multi-source sensor data integrated by the riding protection equipment, and to construct a riding behavior segment sequence through a sliding time window mechanism; The behavior event recognition module is configured to perform time sequence modeling on the behavior segment sequence, construct a behavior graph containing node states and state transition structures, extract action continuity features, and label transition directions. The behavior matching module is configured to perform structural matching between the current behavior graph and a preset inertia skid high-risk behavior template graph, calculate a high-risk matching label of the behavior sequence according to a change range of the node state and a stability degree of the transition path, and perform structural matching between the current behavior graph and a preset inertia skid high-risk behavior template graph. The risk evolution analysis module is configured to generate a current risk evolution path according to the high-risk matching label, and determine whether the current risk evolution path satisfies a set trend change mode. The pre-response control module is configured to execute the pre-response control logic, including starting a pre-inflation instruction of a protection device, adjusting an output rhythm and optical signal parameters of a prompt system, and completing early intervention and closed-loop control on the inertia skid risk.

[0006] Further, the riding data acquisition module comprises: The continuous output signals of acceleration and angular velocity are acquired through the built-in inertial measurement unit, and the speed data and trajectory point information are synchronously acquired through the speed sensor and the positioning module. The acceleration, angular velocity, speed and trajectory point signals are timestamped and aligned in a unified time reference, forming each-dimensional time sequence data streams arranged in the order of sampling time. The duration and step interval of the sliding time window are set, and the continuous data segments are intercepted in the time sequence data streams according to the set sliding rule. Each intercepted data segment is marked as a riding behavior segment, and the original order of the signals in the time dimension within the segment is preserved, forming a riding behavior segment sequence for behavior modeling.

[0007] Further, the behavior event recognition module comprises: Each behavior segment in the behavior segment sequence is defined as a node state, and the node state contains acceleration features, angular velocity features, speed change features and trajectory direction features within the segment. According to the adjacent relationship of the behavior segments in the time sequence, state transition connections are established between the nodes, and each connection reflects the action change relationship between two adjacent segments. The numerical changes between adjacent nodes are analyzed for continuity, the action change is determined to be stable continuation, gradual transition or sudden conversion, and the analysis result is added as a continuity feature to the corresponding state transition connection. According to the trend of the trajectory direction and the angular velocity, the direction attribute of the state transition is determined, and each state transition connection is labeled with a transition direction.

[0008] Further, according to the trajectory direction and the change trend of the angular velocity, the direction attribute of the state transition is determined, and the transition direction is labeled for each state transition connection, including: The change trend of the trajectory direction is calculated from the trajectory point data of the adjacent nodes, and it is judged that the direction change is left deflection, right deflection or straight keeping; The change sign and change amplitude of the angular velocity are extracted from the angular velocity data of the adjacent nodes, and it is judged that the angular velocity change is increasing, decreasing or keeping stable; The trajectory direction change trend and the angular velocity change result are combined and mapped to generate corresponding direction attribute categories, including left turning, right turning, straight keeping, left accelerating turning, right accelerating turning, etc.; The generated direction attribute categories are assigned to the corresponding state transition connections as the transition direction labels of the connections.

[0009] Further, the behavior matching module includes: A preset inertia side-slip high-risk behavior template graph is called, the inertia side-slip high-risk behavior template graph contains node states and their connection relationships arranged in time sequence, and each node state corresponds to a reference range of the feature value of each sensor in the inertia side-slip process; The node states in the current behavior graph are matched one by one with the node states in the template graph in time sequence, and the feature difference value of each matched node is calculated; The connection relationship between the matched nodes is checked for structural consistency, it is judged whether the state transition path is consistent with the connection path of the template graph, and the stability degree of the transition path is evaluated; When the node feature difference value is within the allowable range set by the template graph and the stability degree of the transition path reaches the set threshold, the behavior graph is marked as a high-risk matching state, a high-risk matching label is generated, and it is attached to the current behavior sequence.

[0010] Further, the specific calculation process of the feature difference value and the structural consistency check includes: The calculation process of the feature difference value is: the numerical interval of each sensor feature in the node state of the current behavior graph and the reference numerical interval of the corresponding node state of the template graph are obtained respectively, the absolute value of the difference of each feature in the numerical size and the consistency of the change direction are calculated, it is judged whether the absolute value of the difference is less than the preset allowable fluctuation amplitude, and when all features meet the allowable fluctuation amplitude, the node is determined as a feature matching node; The process of structural consistency check is: the connection path between the nodes in the behavior graph is tracked in time sequence, the direction attribute and continuity identifier of each transition connection in the path are recorded, and the direction change mode of the path is compared with the direction change mode in the template graph one by one, if the direction attribute and continuity identifier of each transition connection are consistent with the template Figure OneIf the feature difference value judgment and the structure consistency check are both in accordance with the set conditions, the corresponding node and path combination are marked as a structure matching unit. When the feature difference value judgment and the structure consistency check are both in accordance with the set conditions, the corresponding node and path combination are marked as a structure matching unit.

[0011] Further, the risk evolution analysis module comprises: The nodes and adjacent nodes marked with high-risk matching labels are sequentially arranged in time sequence to form a current risk evolution path, and the path comprises sensor feature values, direction attributes and state transition relationships of each node. The trend of the feature values of each node in the risk evolution path is analyzed to determine whether the change direction and change amplitude of the acceleration, angular velocity, speed and trajectory direction meet the preset trend change mode requirements. During the trend change mode determination process, the feature changes of the continuous nodes in the path are accumulated and the time is detected to determine whether the duration of the trend feature meets the set minimum duration threshold. When the trend change mode meets the preset requirements, and the duration of the trend feature meets the minimum duration threshold, and the sequence of the state transition direction in the path is consistent with the preset direction change mode, it is determined that the current risk evolution path enters the determination interval. Trigger the pre-response control logic to send an execute pre-response instruction to the protection control process.

[0012] Further, the pre-response control module comprises: Receive the pre-response instruction and parse the pre-response instruction into corresponding protection device control parameters, prompt system control parameters and optical signal control parameters. Send a pre-inflation control instruction to the riding protective equipment, and continuously monitor the airbag pressure and inflation rate. Adjust the output rhythm of the prompt system, shorten the interval time of the sound prompt or vibration prompt to within the set value, and add inertia side slip risk prompt information to the prompt content to improve the attention of the rider. Modify the control parameters of the optical signal, including increasing the flashing frequency of the light source, increasing the light intensity or changing the color of the light source. During the execution of the pre-response control action, continuously collect the working state data of the protection device and the prompt system, and compare and analyze with the risk state change result.

[0013] The technical effects and advantages of the intelligent perception control system for riding protective equipment of the present application are: The application realizes continuous monitoring, accurate identification and early intervention of high-risk riding behaviors such as inertial side-slip by constructing a closed-loop intelligent perception control system integrating riding data collection, behavior event identification, behavior matching, risk evolution analysis and pre-response control. The system constructs a riding behavior graph with time continuity based on multiple sensors, and performs feature difference value calculation and structural consistency check with the high-risk behavior template graph of inertial side-slip, so as to trigger pre-response control when the risk is still in the early stage of evolution. Compared with the traditional protection mode which only relies on single sensor threshold triggering, the application can significantly reduce the false positive rate and the false negative rate, realize the advance control of risk response, avoid the problems of hysteresis and insufficient protection caused by starting the protection device after the risk is fully revealed, and improve the overall protection timeliness and intervention accuracy.

[0014] In addition, the application introduces a feature difference value judgment mechanism, a state transition path structural consistency judgment mechanism and a risk trend change mode comparison mechanism, which ensures that risk identification is not only based on single-point abnormal value, but also combines the dynamic evolution relationship and direction attribute sequence of multi-dimensional features, so as to realize the stability and traceability of risk judgment. Through the risk level control parameter comparison table, the pre-response control module can convert the risk identification result into an executable parameter set containing the pre-inflation target pressure, the inflation rate range, the prompt rhythm and the optical signal adjustment scheme, and real-time monitor the matching degree of control action and risk state during execution, supporting abnormal correction and closed-loop recording. This cooperative mechanism enables the system to complete early protection and prompting under the condition of minimum intervention, avoids unnecessary frequent triggering to disturb the rider, and improves the service life of the protection device and the self-adaptability of the system in complex road environment. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 The application is a structural schematic diagram of an intelligent perception control system for riding protection equipment. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0017] To achieve the above-mentioned purpose, Figure 1 The application is a structural schematic diagram of an intelligent perception control system for riding protection equipment, specifically including a riding data collection module, a behavior event identification module, a behavior matching module, a risk evolution analysis module and a pre-response control module, and the modules are connected through signals. The riding data acquisition module is configured to acquire time sequence signals containing acceleration, angular velocity, speed and trajectory points according to multi-source sensor data integrated in the riding protective equipment, and to construct a riding behavior segment sequence through a sliding time window mechanism. The behavior event recognition module is configured to model the behavior segment sequence in time sequence, to construct a behavior graph containing node states and state transition structures, to extract action continuity features and to label transition directions. The behavior matching module is configured to perform structural matching between a current behavior graph and a preset high-risk behavior template graph of inertial side-slip, to calculate a high-risk matching label of the behavior sequence according to a change amplitude of the node state and a stability degree of the transition path, and to perform structural matching between a current behavior graph and a preset high-risk behavior template graph of inertial side-slip. The risk evolution analysis module is configured to generate a current risk evolution path according to the high-risk matching label, to judge whether the current risk evolution path satisfies a set trend change mode, and to trigger a pre-response control logic when the trend feature continuously exists and satisfies an entering judgment interval condition. The pre-response control module is configured to execute the pre-response control logic, including starting a pre-inflation instruction of the protective device and adjusting output rhythm and optical signal parameters of the prompt system, to complete early intervention and closed-loop control of the inertial side-slip risk.

[0018] The riding data acquisition module is configured to acquire complete and structured riding behavior basic data by collaboratively calling multi-source sensors integrated in the equipment, and to construct a riding behavior segment sequence with time continuity, including the following implementation steps: Acceleration signals and angular velocity signals generated in the riding process are acquired in real time by an inertial measurement unit (IMU) built in the riding protective equipment, wherein the acceleration signals are used to reflect linear motion changes of the rider on three-dimensional coordinate axes, and the angular velocity signals are used to reflect rotation, tilting or posture change processes of the rider. Both of the above-mentioned signals are continuously output and acquired according to a fixed sampling period, for example, the sampling period can be set to collect once every 10 milliseconds to ensure that the data has sufficient time resolution. By equipping the speed sensor and the positioning module, the speed data and the trajectory point information in the riding process are synchronously acquired. The speed data reflects the current motion rate, and the trajectory point information is used to represent the current spatial position coordinates, which is usually realized by the GPS or GNSS mode. The positioning module should have the differential enhancement or auxiliary positioning capability to improve the integrity of the position information in the shielding environment. A unified time reference is adopted to perform the timestamp alignment processing on the acceleration, angular velocity, speed and trajectory point signals. The time reference can be provided by the master processor in the equipment. The master processor allocates a unified timestamp for all data sampling processes through the synchronous system clock, and arranges different modal data into a unified sampling sequence according to the timestamp, to form each-dimensional time sequence data stream with a clear time sequence. In actual implementation, if there is a missing modal signal, the latest valid value in the previous sampling period can be filled to ensure the complete continuity of the data time sequence. A sliding time window is set to extract the behavior segment data from the time sequence data stream. The duration of the sliding time window should be set according to the response speed of the actual riding behavior change, for example, it can be set to 1 second. The step interval is the time step length of the sliding window moving forward each time, for example, it is set to 200 milliseconds.

[0019] For example, in a riding process of entering a left turn on a city road downhill, the system collects acceleration signals and angular velocity signals every ten milliseconds with a unified time reference, and synchronously acquires speed data and trajectory point information from the speed sensor and the positioning module. The sliding time window is set to have a duration of 1 second and a step interval of 200 milliseconds. According to the sliding rule, continuous data segments are sequentially intercepted from each-dimensional time sequence data stream, each data segment is marked as a riding behavior segment and the original time sequence of all sampling points in the segment is preserved, to form a sequence of riding behavior segments for behavior modeling. When the positioning signal is temporarily missing at the entrance of the tunnel, the timestamp alignment is performed according to the rule of filling the latest valid value in the previous sampling period, to ensure the continuity of each-dimensional time sequence data stream. The valid value refers to the sampling point that meets the conditions of complete timestamp, sensor self-checking, and value falling within the physical range of the sensor and the environmental safety range. The sample that does not meet any condition is recorded as an invalid sample and is not used for filling.

[0020] It should be noted that the sliding time window continuously advances according to the set step interval in the application process. In each window period, a complete data segment containing acceleration, angular velocity, speed and trajectory point is intercepted from the time sequence data stream, as an independent behavior data block. For each data segment extracted from the sliding time window, the system labels it as a riding behavior segment, and in the labeling process, the original time sequence and corresponding signal data of all sampling points in the data segment are preserved to ensure that each segment has internal consistency in time sequence, and all generated behavior segments are arranged in sequence to form a structured riding behavior segment sequence with time evolution characteristics.

[0021] The behavior event recognition module specifically includes the following implementation steps: Each behavior segment in the behavior segment sequence is defined as a node state, which is composed of four types of data: acceleration features, angular velocity features, speed change features, and trajectory direction features within the segment. Acceleration features are used to describe the changes in linear motion acceleration within the segment. By collecting and arranging the three-axis acceleration data of each sampling point in the segment in time sequence, the maximum acceleration value, minimum acceleration value, and acceleration change range within the segment are calculated. Angular velocity features are used to reflect the dynamic characteristics of rotational motion within the segment. They are obtained by collecting and analyzing the three-axis angular velocity signals of the gyroscope. Speed change features are obtained by comparing the linear speed values at the start and end times of the segment and combining the fluctuation of the speed curve at the intermediate sampling points, reflecting the acceleration and deceleration state and change amplitude of the speed. Trajectory direction features are obtained by projecting the positioning data onto the horizontal plane and analyzing the overall direction offset and direction stability of the motion trajectory within the segment.

[0022] After defining the node state, state transition connections are established between nodes according to the adjacent relationship of behavior segments in the time sequence. Each connection reflects the action change relationship between two adjacent segments. The connection is bound to the corresponding start node and target node identifier when generated, and then the numerical changes between adjacent nodes are analyzed for continuity. This involves comparing the four types of feature values corresponding to the two nodes one by one, calculating the change amplitude and change rate of each type of feature. If all feature change amplitudes are within the preset stable interval, it is determined to be stable continuation. If the overall feature change amplitude is small but the change rate shows a slow trend, it is determined to be a gradual transition. If the change amplitude of one or more types of features exceeds the mutation threshold and the change rate is significantly higher than the normal interval, it is determined to be a sudden transition, and the continuity analysis result is attached to the corresponding state transition connection as a continuity feature.

[0023] On this basis, according to the change trend of the trajectory direction and the angular velocity, the direction attribute of the state transition is determined, that is, whether the motion direction of adjacent segments is consistent, turns or reverses is judged, and combined with the positive and negative trend and the amplitude size of the angular velocity change, it is marked as straight stable, left turn, right turn or reverse turn and other clear direction types, ensuring that each state transition connection has direction marking, thereby forming a complete behavior graph containing node state, feature change information and direction attribute, providing basic data structure for subsequent behavior pattern analysis and risk judgment.

[0024] According to the change trend of the trajectory direction and the angular velocity, the direction attribute of the state transition is determined, and the transition direction is marked for each state transition connection, the specific steps are: When it is necessary to determine the direction attribute of the state transition according to the change trend of the trajectory direction and the angular velocity, the system first acquires the trajectory point data of adjacent nodes, the trajectory point data is a sequence of continuous spatial coordinate points collected at a uniform time interval; by calculating the spatial coordinate difference between each adjacent node pair in turn, the instantaneous direction vector of each trajectory is obtained, and then the included angle size and the positive and negative signs between adjacent direction vectors are compared to judge the change trend of the trajectory direction belongs to one of the three types of left deflection, right deflection or straight keeping, wherein, left deflection means that the current trajectory direction has counterclockwise rotation relative to the direction vector of the previous segment, right deflection means that the current trajectory direction has clockwise rotation relative to the direction vector of the previous segment, and straight keeping means that the included angle change is close to zero and within a preset straight keeping tolerance range.

[0025] After completing the trajectory direction trend judgment, the angular velocity data of adjacent nodes is extracted from the inertial measurement unit output corresponding to the trajectory point data, the angular velocity data is the instantaneous angular velocity value measured at the same time stamp, the unit is degree per second, in this example, the angular velocity is measured in degrees per second, if measured in radians per second, it is converted by a fixed proportion.

[0026] By calculating the angular velocity difference of adjacent nodes and comparing its sign and the value change amplitude, the angular velocity change is judged to be one of the three types of increase, decrease or keep stable. When the angular velocity change is increased, it means that the current turning speed is rising; when the angular velocity change is decreased, it means that the turning speed is falling; when it is kept stable, it means that the turning speed change is within a preset stable threshold range; Subsequently, the trajectory direction change trend and the angular velocity change result are combined and mapped to generate the corresponding direction attribute category, the rules of this combination and mapping are: When the trajectory direction is left deflection and the angular velocity increases, the direction attribute category is left acceleration steering; when the trajectory direction is left deflection and the angular velocity decreases, the direction attribute category is left deceleration steering; when the trajectory direction is right deflection and the angular velocity increases, the direction attribute category is right acceleration steering; when the trajectory direction is right deflection and the angular velocity decreases, the direction attribute category is right deceleration steering; when the trajectory direction is straight keeping and the angular velocity changes stably, the direction attribute category is straight keeping.

[0027] According to different combination relationships, the direction attribute category can also include, but is not limited to, the defined types such as left steering, right steering, straight keeping, left acceleration steering, right acceleration steering, etc. All generated direction attribute categories will be given the state transition connection corresponding thereto, as the transition direction label information of the connection, and stored in the state transition data structure, for subsequent path prediction, risk identification and control decision calling.

[0028] The behavior matching module specifically includes the following implementation steps: Firstly, the high-risk behavior template graph of inertia skidding stored in the system behavior template library is called, and the template graph is composed of multiple node states arranged in time sequence and the connection relationship thereof. The node state is used to describe the characteristic value range of various sensors (including acceleration sensor, angular velocity sensor, attitude sensor and position sensor, etc.) in the process of inertia skidding. These characteristic value ranges are obtained by statistical analysis of a large number of inertia skidding experimental sample data in the system development stage, and are determined after expert verification. Each node state in the template graph is attached with time stamp information and connection path information between the previous and next node states, and the connection path means the order and logical constraint relationship of the behavior transition from one node state to another node state; The node states arranged in time sequence in the current behavior graph are matched one by one with the corresponding node states in the high-risk behavior template graph of inertia skidding. In the matching process, the system extracts the sensor characteristic values of the current node state, and compares each item with the reference characteristic value range set in the template node. The characteristic difference value of each matched node is calculated. The calculation of the characteristic difference value adopts an absolute difference value mode, that is, the current node sensor characteristic value is subjected to difference operation with the template node reference value, and the absolute value of the obtained result is taken as the difference measure of the characteristic item.

[0029] For a multi-dimensional feature combination node, the difference values of each feature item are calculated and compared with the respective allowed fluctuation amplitude one by one. When the absolute amount of the difference values of all features are less than the respective allowed fluctuation amplitude, and the consistency of the change direction of all features is consistent, the node is determined as a feature matching node.

[0030] Then, the structural consistency of the connection between the matching nodes is checked, that is, whether the state transition path in the current behavior graph is completely consistent with the connection path of the template graph, and the stability of the transition path is evaluated. The stability evaluation is based on the smoothness of the sensor feature value change in the state transition process. If the change rate is within the preset fluctuation threshold and the continuous time period reaches the minimum duration requirement, the transition path is considered stable. After the above comparison and check are completed, if the comprehensive feature difference value of all matching nodes is within the allowable range set by the template graph, and the stability of the transition path reaches the set stability threshold, the system will mark the current behavior graph as a high-risk matching state and generate a corresponding high-risk matching label. The high-risk matching label contains information such as matching time, trigger node number, key feature item, and difference value details, which is used for subsequent risk response module for linkage control. The generated high-risk matching label will be attached to the current behavior sequence to ensure that the subsequent control logic can accurately identify the high-risk state and execute the predetermined protection strategy when receiving the behavior sequence.

[0031] For example, the behavior matching module detects and matches the high-risk behavior of inertia skidding in an actual scenario of motorcycle riding. Specifically, the rider drives the motorcycle at a speed of about 50 kilometers per hour into a left-turn wet road. The riding data collection module collects continuous signals through the built-in inertial measurement unit at the initial stage of entering the curve, including lateral acceleration of 0.85 meters per second, angular velocity of 5° / second, speed change characteristic representative value of 0, and trajectory direction consistent with the road center line; at the middle stage of entering the curve, the lateral acceleration increases to 1.45 meters per second, the angular velocity increases to 18° / second, the speed decreases by 3 kilometers per hour, and the trajectory direction is left biased; at the later stage of the curve, the lateral acceleration fluctuates and rises to 2.1 meters per second, the angular velocity jumps to 26° / second, the speed continues to decrease by 5 kilometers per hour, and the trajectory direction forms an angle of 18° with the road tangent; The three pieces of time sequence data are respectively constructed as node states, and the state transition paths between the nodes are established in time sequence, and the paths are in turn: straight stable to left acceleration steering to left steering. The behavior matching module calls a preset high-risk behavior template graph of inertial side-slip, and the reference feature interval recorded in the template graph is: the lateral acceleration continuously increases and is in the interval of 1.4 to 2.5 m / s², the angular velocity is in the interval of 15° / s to 30° / s and continuously rises, the speed decreases by 3 to 8 km / h in a short time, the trajectory direction cumulative deflection angle exceeds 15°, and the path direction mode is straight stable to left acceleration steering to left steering; the behavior matching module matches each node state of the current behavior graph and the template node state one by one according to the order of the template nodes, calculates the numerical difference absolute amount of the lateral acceleration, the angular velocity, the speed change feature representative value and the trajectory deflection angle, judges the consistency of the change direction of each feature, and confirms that the difference absolute amount is less than the preset allowable fluctuation amplitude. In this embodiment, the lateral acceleration difference between node 1 and the first node of the template is 0.05 m / s², the angular velocity difference is 0° / s, the lateral acceleration and angular velocity difference between node 2 and the middle node of the template are both 0, and the feature difference value between node 3 and the last node of the template also meets the range requirement. Further, the behavior matching module checks the structural consistency of the state transition paths between the nodes, confirms that the direction attribute sequence is consistent with the template and that there is no interruption or reverse transition in the path. Finally, the system marks the behavior graph as a high-risk matching state, generates a high-risk label of inertial side-slip and attaches it to the current behavior sequence.

[0032] The specific calculation process of the feature difference value and the structural consistency check includes: The calculation process of the feature difference value is performed according to the following steps: first, the numerical interval of each sensor feature in the node state of the current behavior graph and the reference numerical interval of the corresponding node state of the template graph are obtained respectively, wherein the reference numerical interval is a closed interval composed of the lower limit value and the upper limit value obtained by statistical analysis of the inertial side-slip samples in the sample collection stage, and is fixed in the template graph through the calibration process; For each sensor feature, the relative position of the current feature representative value to the reference numerical interval is taken as the difference measurement basis: when the current feature representative value falls within the reference numerical interval, the difference absolute amount of the feature is recorded as zero; When the current feature representative value is lower than the reference lower limit, the distance between the current feature representative value and the reference lower limit is taken as the difference absolute amount of the feature; When the current feature representative value is higher than the reference upper limit, the distance between the current feature representative value and the reference upper limit is taken as the difference absolute amount of the feature. The current feature representative value is obtained by taking the median value of all sampling points of the same feature in the segment corresponding to the node state after denoising processing, so as to avoid the influence of extreme values.

[0033] The system simultaneously determines the consistency of the change direction of each feature: the smooth representative values of the feature at the start time and end time of the node state segment are compared in sequence, and if the end representative value is greater than the start representative value and the template node records the upward trend for the feature, the direction is consistent; If the end representative value is less than the start representative value and the template node records the downward trend, the direction is consistent. If the difference between the two does not exceed the minimum significant change and the template node records the stable trend, the direction is consistent. The minimum significant change is set as a fixed threshold according to the comprehensive level of sensor quantization noise and environmental noise during factory calibration. The system pre-sets the allowable fluctuation range for each feature, which is the maximum acceptable distance between the current feature representative value and the reference interval boundary in normal non-risk samples. The allowable fluctuation range is obtained by statistics and stored in the parameter table. When the absolute value of the difference of all features is less than the respective allowable fluctuation range, and the consistency of the change direction of all features is consistent, the node is determined as a feature matching node. If any feature does not meet any of the two conditions, the node is not determined as a feature matching node and the mismatch reason is marked in the record for tracing.

[0034] The process of structure consistency check traces the connection path between nodes in the behavior graph in chronological order and compares the path attributes and template graph records one by one. First, read the candidate path segment of the same length as the template graph in the current behavior graph, and record the direction attribute and continuity identifier of each transition connection in the candidate path segment. The direction attribute is the category label generated based on the trajectory direction and angular velocity trend, and the continuity identifier is one of stable continuation, gradual transition or sudden change. The direction attribute sequence of the candidate path segment is compared with the direction attribute sequence of the corresponding path of the template graph in a one-to-one correspondence. If the direction attributes are completely consistent at all corresponding positions, the direction attribute verification is passed. Otherwise, the direction attribute verification fails and the consistency check is aborted. Then, the system compares the continuity identifier sequence of the candidate path segment with the continuity identifier sequence of the corresponding path of the template graph item by item. If the continuity identifiers of all corresponding positions are consistent, and there is no missing transition connection between any nodes in the candidate path segment, it is determined that there is no interruption in the path. At the same time, check if there is a reverse time sequence, i.e. whether the timestamp of any transition connection is earlier than the timestamp of its start node or later than the timestamp of its target node. If such a situation exists, it is determined as a reverse transition and as inconsistent. Only when the direction attribute verification is passed, the continuity identifier verification is passed, and no interruptions or reverse transitions are found, the path structure consistency is determined to meet the requirements. When the feature difference value judgment and the structural consistency check satisfy the set conditions at the same time, the system marks the combination of the corresponding feature matching node and its connected path segment as a structural matching unit, and generates a structural matching unit record in the behavior graph, which contains the matching range, the matching time, the passed check type, and the empty non-passing items.

[0035] The risk evolution analysis module specifically includes the following implementation steps: The nodes marked with high-risk matching labels by the behavior matching module are arranged in time sequence, and the immediately preceding and following nodes are included in the same continuous path. If necessary, the expansion can continue forward and backward according to the preset maximum path coverage duration until the minimum path length requirement for judgment is met, thereby forming the current risk evolution path.

[0036] The current risk evolution path stores a node list in time sequence, each node containing sensor feature values (acceleration feature representative value, angular velocity feature representative value, speed change feature representative value, and trajectory direction feature representative value), direction attribute, and state transition relationship between adjacent nodes.

[0037] Meanwhile, to ensure the comparability of different modal features, the sensor feature values are taken as the median values after denoising processing in each node segment; the direction attribute is the discrete category obtained based on the trajectory direction and angular velocity change trend; and the state transition relationship includes the transition direction from the starting node to the target node and the continuity feature identifier.

[0038] The trend analysis of the feature values of each node in the path is performed by comparing the same feature representative values of adjacent nodes in sequence, determining whether they increase, decrease, or remain stable from the previous node to the next node, and distinguishing between stable and non-stable changes according to the preset minimum significant change amount; at the same time, the change amplitude is divided into three categories: small, medium, and large, according to the preset amplitude classification boundary. The trend change mode is a set of rules extracted and solidified by the system during sample training, which clearly specifies the joint evolution relationship of each feature when inertia skidding risk occurs, for example: continuous increase in angular velocity and reaching medium amplitude or above, continuous deflection of trajectory direction and cumulative increase in deflection angle, and change of speed change feature from decrease to stability or small amplitude fluctuation in a short time.

[0039] The system checks whether the above joint evolution relationship is established simultaneously on each node on the path to determine whether the change direction and amplitude of features such as acceleration, angular velocity, speed, and trajectory direction meet the preset trend change mode requirements.

[0040] After completing the trend change mode determination, the risk evolution analysis module performs cumulative time detection on the continuous node segments in the path that meet the trend change mode: The duration of the continuous section is calculated with the timestamps of the first node and the last node of the section, and compared with a preset minimum duration threshold. The setting of the minimum duration threshold is determined according to the statistical results of the duration difference between normal riding and confirmed inertia skidding events in sample data, and is fixed as a parameter at the factory calibration of the device. If the duration reaches or exceeds the minimum duration threshold, and the state transition direction sequence corresponding to the continuous section is compared with the preset direction change pattern one by one, the direction change pattern is a finite sequence obtained by sample induction; For example, "straight stable to left acceleration steering and then to left steering" or "straight stable to right acceleration steering and then to right steering", when the comparison result shows that they are completely consistent in order and category, and there is no time sequence interruption or reverse transition in the path, the module determines that the current risk evolution path enters the judgment interval. After entering the judgment interval, the risk evolution analysis module immediately triggers the pre-response control logic and sends an execution pre-response instruction to the protection control process; the execution pre-response instruction carries the pre-inflation control parameters, prompt system control parameters and optical signal control parameters in the form of structured parameters, and the protection control process executes the subsequent pre-inflation, prompt rhythm adjustment and light signal adjustment operations according to the parameter content after receiving the instruction, so as to complete the early intervention before the inertia skidding risk fully appears and lay the data and timing foundation for closed-loop control.

[0041] The pre-response control module specifically includes the following implementation steps: After receiving the pre-response instruction sent by the risk evolution analysis module, the instruction is first parsed, and the control intention carried therein is mapped into three types of parameter sets, namely protection device control parameters, prompt system control parameters and optical signal control parameters.

[0042] The protection device control parameters at least include the pre-inflation target pressure, the maximum inflation duration, the allowed inflation rate range and the safe shutdown condition; the above parameters are derived from the factory calibration table and the risk level-control parameter correspondence table, and the risk level-control parameter correspondence table is established according to the correspondence between different risk levels and expected response advance in the sample training stage, and is written into the storage area at device initialization; The prompt system control parameters at least include the prompt mode, the output rhythm and the prompt duration, wherein the prompt mode is selected between sound prompt and vibration prompt, the output rhythm is represented by the time interval between adjacent two prompts, and the set value is given by the risk level-control parameter correspondence table, and the prompt content is set as the inertia skidding risk prompt information when executed, so that the rider can intuitively identify the current risk situation; The optical signal control parameters include at least flicker frequency, light intensity and light source color. The flicker frequency is represented by the number of times of turning on and off the light source per unit time, the light intensity is represented by the rated brightness level of the light source, and the light source color is represented by a value in a fixed color set supported by the device. The above three parameters are written into the register configuration area of the optical signal driving unit after the instruction is parsed.

[0043] After the parameter parsing is completed, a pre-inflation control instruction is sent to the riding protective equipment according to the protective device control parameters, the inflation mechanism is started, and the sensor output of the airbag pressure and the inflation rate is continuously read at a fixed sampling period. The instantaneous pressure and the instantaneous inflation rate obtained in each sampling period are compared with the pre-inflation target pressure and the allowed inflation rate range one by one: when the instantaneous pressure is not lower than the pre-inflation target pressure in the continuous detection period, or when the maximum inflation duration reaches, the inflation stop logic is triggered; when the instantaneous inflation rate continuously exceeds the allowed inflation rate range, the safety shutdown condition is triggered and the abnormal reason code is recorded; At the same time, the pre-response control module adjusts the output rhythm of the prompting system according to the prompting system control parameters, sets the interval time of the sound prompt or the vibration prompt to a fixed interval not greater than a set value, and adds inertia side slip risk prompt information in the prompt content; according to the optical signal control parameters, the working state of the optical signal is modified, the flicker frequency is set to the value specified by the instruction, the light intensity is switched to the specified level, and the light source color is switched to the specified color.

[0044] During the execution of the pre-response control action, the pre-response control module continuously collects the working state data of the protective device and the prompting system, and compares and analyzes the risk state change result. The working state data at least includes continuous readings of airbag pressure, continuous readings of inflation mechanism working current, instantaneous inflation rate calculation result, actual trigger time sequence of sound prompt or vibration prompt, actual on-off time sequence of optical signal and brightness level record; The risk state change result is output by the risk evolution analysis module according to a unified time reference, and at least contains the risk level label and the state transition direction sequence in the continuous period. The comparison and analysis is performed in the time stamp alignment mode: the working state data and the risk state change result in the same period are compared and analyzed period by period, to determine whether the pre-inflation is completed in the risk level rising stage, whether the trigger rhythm of the prompting system and the optical signal is consistent with the output rhythm specified in the risk level-control parameter comparison table, and whether the above control actions are required to exit to standby state after the risk level drops to the basic safety level for a number of continuous detection periods; If the comparison result shows that any control action does not meet the parameter requirement within the corresponding risk stage, the pre-response control module immediately outputs a correction instruction, which is reissued to the corresponding execution unit in the form of parameter override and records the correction time point; if the comparison result shows that all control actions meet the parameter requirement within the corresponding risk stage, the pre-response control module records the pre-response period as a completed state and writes the record back to the event log for subsequent closed-loop analysis and strategy optimization.

[0045] For example, during a night ride, when the inertial side-slip risk value is identified to exceed the preset trigger threshold, the pre-response control module receives a pre-response instruction and parses a modification scheme for the optical signal control parameters: The light source flicker frequency is increased from the original 1 time per second to 3 times per second to enhance the flicker rhythm and enhance the visual attention of the rider and surrounding traffic participants; The light intensity is increased from the original setting of 80 lumens to 120 lumens to enhance the visible distance of the light signal in a low-illumination environment; The light source color is switched from the default white color to the high-alert amber color to form a risk warning signal in the traffic visual system; The adjustment values of these parameters are determined by the system based on road light environment tests and human eye reaction time research during the design stage and stored in the optical signal control parameter library, which are called by the pre-response control module according to the instructions during operation.

[0046] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product.

[0047] Those skilled in the art can realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0048] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0049] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, and all of them should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0050] Finally, the above merely provides the preferred embodiments of the present application, and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. An intelligent sensing control system for ride protective equipment, characterized by: The system comprises a riding data acquisition module, a behavior event identification module, a behavior matching module, a risk evolution analysis module, and a pre-response control module, and the modules are connected through signals; The riding data acquisition module is configured to acquire time sequence signals including acceleration, angular velocity, speed, and trajectory points based on multi-source sensor data of the riding protective equipment set, and to construct a riding behavior segment sequence through a sliding time window mechanism; The behavior event identification module is configured to model the behavior segment sequence in time sequence, construct a behavior graph including node states and state transition structures, extract action continuity features, and label transition directions; The behavior matching module is configured to perform structural matching between a current behavior graph and a preset high-risk behavior template graph of inertial side-slip, calculate a high-risk matching label of the behavior sequence based on a change amplitude of the node state and a stability degree of the transition path, and perform structural matching between a current behavior graph and a preset high-risk behavior template graph of inertial side-slip; The risk evolution analysis module is configured to generate a current risk evolution path based on the high-risk matching label, and determine whether the current risk evolution path meets a set trend change mode, and trigger a pre-response control logic when the trend feature continuously exists and meets an entering judgment interval condition; The pre-response control module is configured to execute the pre-response control logic, including starting a pre-inflation instruction of the protective device, adjusting an output rhythm and optical signal parameters of the prompt system, and completing early intervention and closed-loop control on the inertial side-slip risk.

2. An intelligent sensing control system for ride protective equipment according to claim 1, characterized in that: The riding data acquisition module comprises: The continuous output signals of acceleration and angular velocity are acquired through an inertial measurement unit built in the equipment, and the speed data and trajectory point information are synchronously acquired through a speed sensor and a positioning module; The acceleration, angular velocity, speed, and trajectory point signals are timestamped and aligned based on a unified time reference, and each dimension time sequence data stream arranged in a sampling time sequence is formed; The duration and step interval of the sliding time window are set, and the continuous data segments are intercepted in the time sequence data stream according to the set sliding rule; Each intercepted data segment is marked as a riding behavior segment, and the original order of the signals in the time dimension is retained, and a riding behavior segment sequence for behavior modeling is formed.

3. An intelligent sensing control system for ride protective equipment as defined in claim 2, wherein: The behavior event identification module comprises: Each behavior segment in the behavior segment sequence is defined as a node state, and the node state includes acceleration features, angular velocity features, speed change features, and trajectory direction features in the segment; State transition connections are established between the nodes according to the adjacent relationship of the behavior segments in the time sequence, and each connection reflects the action change relationship between two adjacent segments; The numerical change between adjacent nodes is analyzed for continuity, and it is determined whether the action change is stable continuation, gradual transition, or sudden conversion, and the analysis result is added as a continuity feature to the corresponding state transition connection; The direction attribute of the state transition is determined according to the trajectory direction and the change trend of the angular velocity, and the transition direction is labeled for each state transition connection.

4. An intelligent sensing control system for ride protective equipment as defined in claim 3, wherein: The direction attribute of the state transition is determined according to the trajectory direction and the change trend of the angular velocity, and the transition direction is labeled for each state transition connection, including: The change trend of the trajectory direction is calculated from the trajectory point data of adjacent nodes, and it is determined whether the direction change is left deflection, right deflection, or straight keeping; Extract the change sign and change amplitude of the angular velocity from the angular velocity data of the adjacent nodes, and judge whether the angular velocity changes increase, decrease or remain stable; Combine the trajectory direction change trend with the angular velocity change result to generate the corresponding direction attribute category, including left turn, right turn, straight line stability, left acceleration turn, right acceleration turn; Assign the generated direction attribute category to the corresponding state transition connection as the transition direction label of the connection.

5. An intelligent sensing control system for ride protective equipment as defined in claim 4, wherein: The behavior matching module includes: Call the preset high-risk inertia side-slip behavior template graph, which contains node states and their connection relationships arranged in time sequence, and each node state corresponds to the reference range of each sensor feature value in the inertia side-slip process; Match the node states in the current behavior graph with the node states in the template graph in time sequence one by one, and calculate the feature difference value of each matched node; Check the structural consistency of the connection relationship between the matched nodes, judge whether the state transition path is consistent with the connection path of the template graph, and evaluate the stability degree of the transition path; When the node feature difference value is within the allowable range set by the template graph and the stability degree of the transition path reaches the set threshold, mark the behavior graph as a high-risk matching state, generate a high-risk matching label, and attach it to the current behavior sequence.

6. An intelligent sensing control system for ride protective equipment as defined in claim 5, wherein: The specific calculation process of the feature difference value and the structural consistency check includes: The calculation process of the feature difference value is: obtain the numerical interval of each sensor feature in the current behavior graph node state and the reference numerical interval of the corresponding node state in the template graph, calculate the absolute value of the difference and the consistency of the change direction of each feature in the numerical size, judge whether the absolute value of the difference is less than the preset allowable fluctuation amplitude, and when all features meet the allowable fluctuation amplitude, determine the node as a feature matching node; The process of structural consistency check is: track the connection path between the nodes in the behavior graph in time sequence, record the direction attribute and continuity identifier of each transition connection in the path, and compare the direction change mode of the path with the direction change mode in the template graph one by one. If the direction attribute and continuity identifier of each transition connection are consistent with the template graph, and there is no interruption or reverse transition in the path, it is determined that the path structure consistency meets the requirements; When the feature difference value determination and the structural consistency check both meet the set conditions, the corresponding node and path combination is marked as a structure matching unit.

7. An intelligent sensing control system for ride protective equipment as defined in claim 6, wherein: The risk evolution analysis module includes: Arrange the nodes and adjacent nodes marked with high-risk matching labels in time sequence to form a current risk evolution path, which contains the sensor feature values, direction attributes and state transition relationships of each node in the path; Analyze the trend of the feature values of each node in the risk evolution path to determine whether the change direction and change amplitude of the acceleration, angular velocity, speed and trajectory direction features meet the preset trend change mode requirements; In the trend change mode determination process, accumulate the time of the feature change of the consecutive nodes in the path to determine whether the duration of the trend feature reaches the set minimum duration threshold; When the trend change pattern meets the preset requirements, the duration of the trend feature reaches the minimum duration threshold, and the state transition direction sequence in the path is consistent with the preset direction change pattern, it is determined that the current risk evolution path enters the determination interval; Triggering the pre-response control logic, sending the execution pre-response instruction to the protection control process.

8. An intelligent sensing control system for ride protective equipment according to claim 7, wherein: The pre-response control module includes: Receiving the pre-response instruction and parsing the pre-response instruction into corresponding protection device control parameters, prompt system control parameters and optical signal control parameters; Sending the pre-inflation control instruction to the riding protective equipment and continuously monitoring the airbag pressure and inflation rate; Adjusting the output rhythm of the prompt system, shortening the interval time of the sound prompt or vibration prompt to within the set value, and adding inertia side slip risk prompt information to the prompt content to improve the attention of the rider; Modifying the control parameters of the optical signal, including increasing the flashing frequency of the light source, increasing the light intensity or changing the color of the light source; Continuously collecting the working state data of the protection device and the prompt system during the execution of the pre-response control action, and comparing and analyzing with the risk state change result.

Images