Vehicle acceleration event identification method and vehicle

Abstract

The application discloses a vehicle acceleration event identification method and a vehicle, and relates to the technical field of vehicles, wherein the vehicle acceleration event identification method comprises the following steps: acquiring the longitudinal acceleration, the throttle opening, the engine power output parameter and the vehicle mass of the vehicle; determining an acceleration burst impact index based on the longitudinal acceleration and the throttle opening, the acceleration burst impact index representing the burst strength of the driver's intention to accelerate suddenly; determining an acceleration sustained kinetic energy factor based on the longitudinal acceleration, the engine power output parameter and the vehicle mass, the acceleration sustained kinetic energy factor representing the strength of the sustained power output capability of the vehicle; and identifying an acceleration event based on the acceleration burst impact index and the acceleration sustained kinetic energy factor. The method disclosed by the application can reduce the false positives and false negatives of non-real sudden acceleration, and realizes accurate and continuous identification of real sudden acceleration events.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and in particular to a vehicle acceleration event recognition method and a vehicle. Background Technology

[0002] Accurate identification of vehicle acceleration events is a crucial foundation for driving style evaluation, vehicle status monitoring, and connected vehicle services. Related technologies typically require the timely capture and identification of situations involving rapid vehicle acceleration from driving data collected from in-vehicle terminals or mobile devices.

[0003] To achieve the above objectives, related technologies typically use whether the longitudinal acceleration exceeds a preset fixed threshold as the primary criterion. However, such solutions rely mainly on a single kinematic threshold. In non-realistic rapid acceleration scenarios such as inertial coasting, the vehicle's longitudinal acceleration may be high, but the driver has not made any obvious rapid acceleration operation, and the engine has not provided corresponding power output. In this case, relying solely on the acceleration threshold can easily lead to false alarms. Conversely, when the driver has a clear intention to accelerate rapidly and the engine is continuously outputting large power, but the longitudinal acceleration amplitude does not significantly exceed the threshold due to the large vehicle mass or differences in operating conditions, false alarms are likely to occur.

[0004] Therefore, in application scenarios where vehicles are susceptible to multiple factors and have complex power response characteristics, how to reduce false alarms and false negatives of non-realistic rapid acceleration has become an urgent problem to be solved in related technologies. Summary of the Invention

[0005] This application aims to at least partially address one of the technical problems in related technologies. To this end, the first objective of this application is to propose a vehicle acceleration event recognition method. This method acquires the vehicle's longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass. First, it determines an acceleration burst impact index based on longitudinal acceleration and throttle opening to indicate the intensity of the driver's intent to accelerate rapidly. Then, it determines an acceleration sustaining kinetic energy factor based on longitudinal acceleration, engine power output parameters, and vehicle mass to verify the vehicle's sustained power output capability. Finally, it identifies acceleration events based on the acceleration burst impact index and the acceleration sustaining kinetic energy factor. By jointly using driver operating parameters, kinematic parameters, and engine power output parameters for event recognition, it can eliminate false alarms and missed alarms caused by slope gravity components and inertial coasting when relying solely on acceleration threshold recognition, achieving complete and continuous recognition of genuine rapid acceleration events.

[0006] The second objective of this application is to propose a vehicle.

[0007] To achieve the above objectives, a first aspect of this application proposes a vehicle acceleration event recognition method, the method comprising: acquiring the vehicle's longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass; determining an acceleration burst impact index based on the longitudinal acceleration and the throttle opening, the acceleration burst impact index representing the burst intensity of the driver's intention to accelerate rapidly; determining an acceleration sustained kinetic energy factor based on the longitudinal acceleration, the engine power output parameters, and the vehicle mass, the acceleration sustained kinetic energy factor representing the strength of the vehicle's sustained power output capability; and recognizing the acceleration event based on the acceleration burst impact index and the acceleration sustained kinetic energy factor.

[0008] The vehicle acceleration event recognition method according to embodiments of this application acquires the vehicle's longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass. Based on the longitudinal acceleration and throttle opening, an acceleration burst impact index is determined, representing the intensity of the driver's sudden acceleration intention. Based on the longitudinal acceleration, engine power output parameters, and vehicle mass, an acceleration sustained kinetic energy factor is determined, representing the strength of the vehicle's sustained power output capability. Acceleration events are identified based on the acceleration burst impact index and the acceleration sustained kinetic energy factor. Therefore, this method effectively integrates the driver's operational intention with the vehicle's powertrain response characteristics, overcoming the limitations of the single acceleration threshold method in complex conditions such as slopes and coasting, and significantly improving the accuracy and robustness of rapid acceleration event identification. In other words, when a vehicle is on an uphill section, even with low longitudinal acceleration, if the throttle opening increases rapidly and the engine torque increases simultaneously, the acceleration burst impact index can still reflect the driver's intention to accelerate rapidly. Simultaneously, by combining this with the acceleration sustaining kinetic energy factor calculated from engine torque, engine speed, and vehicle mass, the system can further verify whether the vehicle has the actual power output capability to achieve this acceleration intention, thus avoiding false alarms caused by gravitational interference. Conversely, during coasting or downhill driving, although longitudinal acceleration may briefly increase, the acceleration burst impact index and acceleration sustaining kinetic energy factor are both at low levels due to the near-zero throttle opening and weak engine torque output, allowing the system to exclude false triggers. Therefore, through this dual-dimensional collaborative judgment, accurate and continuous identification of real rapid acceleration events is achieved.

[0009] According to one embodiment of this application, determining the acceleration burst impact index based on the longitudinal acceleration and the throttle opening includes: determining a target longitudinal acceleration based on the longitudinal acceleration; determining the jerk of the target longitudinal acceleration based on the target longitudinal acceleration, wherein the jerk represents the instantaneous rate of change of the target longitudinal acceleration; determining the instantaneous rate of change of the throttle opening based on the throttle opening; and determining the acceleration burst impact index based on the jerk and the instantaneous rate of change of the throttle opening.

[0010] According to one embodiment of this application, determining the acceleration burst impact index based on the jerk and the instantaneous rate of change of the throttle opening includes: calculating the effective value of the jerk within a first preset time period, and calculating the average value of the instantaneous rate of change of the throttle opening within the first preset time period, wherein the effective value is the square root of the average of the squares of the various sampled values ​​of the jerk; when the effective value is greater than a preset jerk threshold and the average value is greater than a preset throttle change rate threshold, determining that the acceleration burst impact index characterizes the start of the rapid acceleration event.

[0011] According to one embodiment of this application, the engine power output parameters include engine torque and engine speed. The step of determining the acceleration sustained kinetic energy factor based on the longitudinal acceleration, the engine power output parameters, and the vehicle mass includes: determining a target longitudinal acceleration based on the longitudinal acceleration; determining engine output power based on the product of the engine torque and the engine speed; determining specific power based on the ratio of the engine output power to the vehicle mass; and weighted summing the ratio of the target longitudinal acceleration to a preset acceleration reference threshold and the ratio of the specific power to a preset specific power reference threshold to obtain the acceleration sustained kinetic energy factor.

[0012] According to one embodiment of this application, determining the target longitudinal acceleration based on the longitudinal acceleration includes: obtaining the longitudinal slope angle of the current road; and using the longitudinal slope angle to perform slope compensation on the longitudinal acceleration to obtain the target longitudinal acceleration.

[0013] According to one embodiment of this application, identifying the acceleration event based on the acceleration burst impact index and the acceleration sustained kinetic energy factor includes: when the acceleration burst impact index represents the start of the rapid acceleration event, determining that the vehicle has started to accelerate and recording the acceleration time; when the vehicle has started to accelerate, if the acceleration sustained kinetic energy factor is greater than a first preset threshold and the target longitudinal acceleration is greater than a preset minimum acceleration threshold, then determining that the vehicle is in an acceleration state and recording the corresponding acceleration parameters.

[0014] According to one embodiment of this application, the method further includes: recording the duration when the acceleration continuous kinetic energy factor is less than the first preset threshold and greater than the second preset threshold; and maintaining the vehicle in an acceleration state if the duration is less than or equal to the preset duration threshold.

[0015] According to one embodiment of this application, after determining that the vehicle is in an acceleration state, the method further includes: acquiring the vehicle speed, and calculating the average value of the acceleration of the target longitudinal acceleration within a second preset time period; and determining that the vehicle exits the acceleration state based on the average value, the throttle opening, and the vehicle speed.

[0016] According to one embodiment of this application, determining that the vehicle exits the acceleration state based on the average value, the throttle opening, and the vehicle speed includes: determining the vehicle speed fluctuation rate based on the vehicle speed, where the vehicle speed fluctuation rate represents the instantaneous change in the vehicle speed as it approaches steady-state cruising; and determining that the vehicle exits the acceleration state when the average value is less than a preset acceleration threshold, the throttle opening is less than a preset cruising throttle opening threshold, and the vehicle speed fluctuation rate is less than a preset vehicle speed fluctuation rate threshold and continues for a preset duration threshold.

[0017] To achieve the above objectives, a vehicle is provided in a second aspect of this application, including a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the above-described vehicle acceleration event recognition method.

[0018] The vehicle according to the embodiments of this application, by executing the above-described vehicle acceleration event identification method, can reduce false alarms and false negatives of non-real rapid acceleration, and achieve accurate and continuous identification of real rapid acceleration events.

[0019] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0020] Figure 1 This is a flowchart of a vehicle acceleration event recognition method according to an embodiment of this application.

[0021] Figure 2 This is a flowchart of a vehicle acceleration event recognition method according to a specific example of this application.

[0022] Figure 3 This is a block diagram of a vehicle according to an embodiment of this application. Detailed Implementation

[0023] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0024] To identify rapid acceleration events, relevant technologies typically employ an acceleration threshold-based judgment method. Specifically, this approach acquires the vehicle's longitudinal acceleration or calculates it from the vehicle speed difference and compares it to a preset threshold. Its basic working principle is that when the acceleration value exceeds a fixed threshold, the corresponding moment is marked as a rapid acceleration event. This approach is widely used primarily because it can mark moments of acceleration exceeding the limit with relatively low data requirements.

[0025] However, this solution performs poorly when applied to scenarios like vehicles, where weight is low and slopes have a significant impact. A fundamental problem is that, in its pursuit of simplicity, the solution's inherent design incorporates acceleration components generated by non-driver-initiated acceleration and fails to verify whether the power source is actually delivering power. Specifically, on long downhill sections, even without accelerating, the acceleration value can easily exceed the rapid acceleration threshold, leading to false alarms; on long uphill sections, even with heavy acceleration, the acceleration value may still be low, resulting in missed alarms. Furthermore, relying solely on acceleration thresholds makes it difficult to distinguish between sudden, rapid acceleration and continuous, steady acceleration, and it cannot maintain the continuity of events during power interruptions during gear shifts, causing a single, continuous rapid acceleration to be fragmented into multiple isolated events.

[0026] In-depth analysis revealed that the root cause of the aforementioned problems lies in the fact that a single acceleration threshold only reflects kinematic results and fails to jointly verify driver operating parameters, vehicle kinematic parameters, and engine power output parameters. Therefore, it cannot effectively distinguish between acceleration changes caused by the driver's active acceleration behavior and passive acceleration fluctuations caused by external factors such as road gradient and inertial coasting. Furthermore, this method lacks the ability to perceive the actual response state of the powertrain, making it difficult to determine whether the vehicle is truly in the power output phase at a specific acceleration level. Therefore, under complex driving conditions, relying solely on acceleration thresholds cannot accurately capture the driver's acceleration intentions, nor can it reliably assess the vehicle's actual acceleration capability, ultimately leading to distorted recognition results.

[0027] To address the aforementioned issues, this application proposes a vehicle acceleration event recognition method. By acquiring the vehicle's longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass, it first determines an acceleration burst impact index based on longitudinal acceleration and throttle opening to indicate the intensity of the driver's intent to accelerate rapidly. Then, it determines an acceleration sustaining kinetic energy factor based on longitudinal acceleration, engine power output parameters, and vehicle mass to verify the vehicle's sustained power output capability. Finally, it identifies acceleration events based on both the acceleration burst impact index and the acceleration sustaining kinetic energy factor. By jointly using driver operating parameters, kinematic parameters, and engine power output parameters for event recognition, it can eliminate false alarms and missed alarms caused by slope gravity components and inertial coasting when relying solely on acceleration thresholds, achieving complete and continuous recognition of genuine rapid acceleration events.

[0028] In one embodiment of this application, the vehicle includes an on-board terminal, a six-axis inertial measurement unit (IMU), a CAN (Controller Area Network) bus interface, a satellite positioning module, and a wheel speed sensor interface. The on-board terminal is connected to the six-axis IMU, the CAN bus, and the wheel speed sensor interface. The six-axis IMU measures the vehicle's longitudinal acceleration and provides attitude information for slope angle calculation. The CAN bus connects to the vehicle's engine control unit (ECU) and is used to acquire throttle opening, engine output torque, and engine speed. The satellite positioning module or wheel speed sensor interface provides a vehicle speed reference. The on-board terminal internally includes a processor, a memory, and an external communication interface. The memory stores programs that can run on the processor. The processor receives data from the six-axis IMU, the CAN bus, and the wheel speed sensor interface, and performs clock synchronization and resampling processing on this data. The external communication interface can upload the identification results to a remote server via a mobile network, or it can simply store the results in local memory.

[0029] In terms of data flow, the six-axis inertial measurement unit outputs longitudinal acceleration data at a sampling frequency of no less than 100 Hz, the engine control unit outputs throttle opening and engine torque data via the CAN bus at a sampling frequency of no less than 50 Hz, and engine speed, vehicle speed, and gradient angle related data are output at a sampling frequency of no less than 20 Hz. All data entering the on-board terminal are timestamped and interpolated and aligned based on the timestamps, ensuring that sampling points from different channels are within the same time series.

[0030] The connection methods and sampling frequencies between the aforementioned system components can be adjusted according to the actual vehicle platform. For example, on models equipped with a factory-installed CAN bus interface, longitudinal acceleration can also be obtained from the vehicle stability control system or electronic control unit; on models without a satellite positioning module, vehicle speed can be calculated solely from wheel speed pulse signals.

[0031] The vehicle acceleration event recognition method and vehicle proposed in this application are described below with reference to the accompanying drawings.

[0032] Figure 1 This is a flowchart of a vehicle acceleration event recognition method according to an embodiment of this application.

[0033] like Figure 1 As shown, the vehicle acceleration event recognition method of this application embodiment may include the following steps: S1 acquires the vehicle's longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass.

[0034] Specifically, during vehicle operation, the onboard terminal first acquires longitudinal acceleration (a_long), throttle opening (throttle_pos), engine power output parameters, and an estimated vehicle mass (mass_est). Longitudinal acceleration generally refers to linear acceleration along the vehicle's forward direction. This quantity can be the original measured acceleration or acceleration compensated for by gravity. Throttle opening generally refers to the opening or equivalent control quantity related to the driver's power request; it can be the mechanical throttle opening or a percentage corresponding to an electronic throttle command. Engine power output parameters may include engine output torque (torque_act) and engine speed (rpm_eng). Engine torque and engine speed refer to a set of parameters characterizing the instantaneous output capability of the powertrain; they can be directly obtained from controller broadcast data or calculated from multiple basic parameters. For example, longitudinal acceleration can be output by a six-axis inertial measurement unit (IMU), throttle opening, engine output torque, and engine speed can be provided by the engine electronic control unit via the CAN interface, and the vehicle mass can be an estimate of the sum of the vehicle's curb weight and the driver's weight, or it can be corrected by combining the load identification results.

[0035] It should be noted that the vehicle mass is not limited to a fixed nominal curb weight. It refers to the mass parameter used to characterize the current force and dynamic response relationship of the vehicle. It can be a static preset value or a dynamic estimate that is updated as the driver, passengers or cargo load change.

[0036] As a specific implementation, the processor can attach a unified timestamp to each data channel and align data from different sampling frequencies to a unified sequence based on the timestamp. For example, longitudinal acceleration data can be sampled at a frequency of at least 100Hz, throttle opening and engine torque at a frequency of at least 50Hz, and engine speed and vehicle speed at a frequency of at least 20Hz. Subsequently, the processor can resample to a unified frequency, such as 100Hz, through linear interpolation, so that sampling points from different sources are at the same time coordinate. This approach reduces the phase shift caused by asynchronous sampling, making subsequent judgments about the timing relationships between acceleration spikes, throttle changes, and power output more stable.

[0037] More generally, the above data can be acquired in a variety of ways. For example, longitudinal acceleration can be obtained not only by direct measurement by the IMU, but also by output from the vehicle stability control system or other electronic control units; vehicle speed can be provided by the positioning module, calculated based on wheel speed pulses by the wheel speed interface, or obtained by a combination of both; throttle opening can be obtained using throttle position sensor signals or corresponding control signals.

[0038] It should be noted that the joint acquisition of multi-source data in step S1 provides a synchronous basis for driver input and vehicle motion response in the identification of the initial stage of rapid acceleration in step S2, and also provides common input of dynamic and kinematic quantities for the judgment of continuous power output capability in step S3. Through this collaboration, problems such as false alarms of ramps, false alarms of inertial coasting, and missed detection of real events that are prone to occur when judging based on a single acceleration threshold are solved.

[0039] S2, based on longitudinal acceleration and throttle opening, determines the acceleration burst impact index, which represents the burst intensity of the driver's intention to accelerate rapidly.

[0040] Specifically, after determining the longitudinal acceleration and throttle opening, the acceleration burst impact index (haoi) can be determined based on these parameters. The acceleration burst impact index represents the intensity of the driver's intent to accelerate rapidly. That is, the acceleration burst impact index is an indicator used to characterize the explosiveness of the initial acceleration, and is not limited to a specific formula. It can be formed based on a product kernel function or through threshold logic output. The burst intensity can be expressed as a binary result indicating whether the acceleration has begun, or as a multi-level intensity or continuous amplitude.

[0041] As a specific implementation, the processor can first obtain the target longitudinal acceleration based on the longitudinal acceleration itself, then differentiate the target longitudinal acceleration to obtain the jerk (jerk_raw), and simultaneously calculate the instantaneous rate of change of the throttle opening (delta_throttle) based on the throttle opening. In terms of time, when the driver rapidly increases the throttle, the vehicle's longitudinal acceleration typically exhibits a pulse-like upward trend, mathematically represented as a jerk spike. If this spike and the instantaneous rate of change of the throttle opening occur synchronously within the same short time window, it can be considered that the driver has a clear intention to accelerate rapidly from a standstill. In other words, this method allows for a more precise pinpointing of the moment of rapid acceleration initiation.

[0042] More generally, the acceleration burst impact index can be represented by a logical quantity, a discrete-level quantity, or a continuous-scoring quantity. For example, it can take the value 1 when the conditions are met and 0 when the conditions are not met; it can also be set with multiple levels to distinguish different intensities of burst; and it can also use a continuous function to fuse jerk and throttle change rate. Anything that can reflect the burst characteristics of the driver applying a power request in a short period of time and triggering a longitudinal response of the vehicle can be included in the scope of the acceleration burst impact index.

[0043] Therefore, the obtained acceleration burst impact index can serve as an event trigger signal, providing a starting basis for the start determination in step S4; however, the start trigger alone is not sufficient to guarantee that the event is a genuine and continuous rapid acceleration, so it is still necessary to combine it with the acceleration continuous kinetic energy factor obtained in step S3 for further verification. The combination of the two can form a closed loop between the sensitive identification at the start moment and the authenticity verification of the subsequent process.

[0044] S3 determines the continuous acceleration kinetic energy factor based on longitudinal acceleration, engine power output parameters, and vehicle mass. The continuous acceleration kinetic energy factor represents the strength of the vehicle's continuous power output capability.

[0045] Specifically, after obtaining longitudinal acceleration, engine power output parameters, and vehicle mass, the acceleration sustaining kinetic energy factor (haki) can be determined based on these parameters. The acceleration sustaining kinetic energy factor represents the strength of the vehicle's sustained power output capability. Unlike simply observing whether the vehicle experiences significant acceleration, this factor considers both the vehicle's current longitudinal motion and the engine's active output capability, thus enabling mechanistic verification of the sustained acceleration state. In other words, the acceleration sustaining kinetic energy factor is an indicator used to describe the vehicle's ability to maintain an active acceleration state over a period of time. When it is dimensionless, it facilitates comparisons between different vehicle models under a unified criterion; however, in other implementations, it can retain physical dimensions, as long as it reflects the normalized overall acceleration level.

[0046] As a specific implementation, the processor can calculate the engine output power using the product of engine torque and engine speed, and then divide the engine output power by the vehicle mass to obtain the specific power sp(t). For example, it can be calculated as sp(t) = (torque_act(t) × rpm_eng(t) / 9549) / mass_est, where torque_act(t) is the engine torque at time t, rpm_eng(t) is the engine speed at time t, and mass_est is the vehicle mass. Subsequently, the processor performs a weighted summation of the ratio of the target longitudinal acceleration to a preset acceleration reference threshold a_base, and the ratio of the specific power to a preset specific power reference threshold sp_base, to obtain the acceleration sustained kinetic energy factor haki(t). In some embodiments, haki(t) = α·(a_true(t) / a_base) + β·(sp(t) / sp_base) can be used, where α and β satisfy α + β = 1, for example, α = 0.6 and β = 0.4; a_base can be 0.8 m / s², and sp_base can be 50 W / kg. By incorporating the acceleration contribution corresponding to the feeling of being pushed back into the seat and the specific power contribution corresponding to the power source output, a comprehensive judgment can be made on whether the vehicle is in a real active acceleration process. Specific power generally refers to the power contribution per unit mass, which can be the ratio of the engine's actual output power to its mass, or the ratio of the equivalent output value after efficiency correction to its mass.

[0047] More generally, power-to-weight ratio can be determined based on engine-side parameters or equivalent wheel-end drive power parameters; the weighting coefficients can also be adjusted according to vehicle type, displacement, vehicle weight range, or driving style calibration results. For example, in some high-powered vehicles, the proportion of β can be appropriately increased to enhance the sensitivity to the engine's active power output characteristics; in other vehicles that emphasize ride comfort, the proportion of α can also be appropriately increased to enhance the representation of actual longitudinal body feel. These approaches can all quantify continuous power output capability.

[0048] Therefore, by continuing to verify the continuous kinetic energy factor after the start, it is possible to avoid misidentifying a vehicle's high acceleration due to downhill gravity or inertial coasting as a driver's active acceleration, thereby improving the reliability of the event recognition results.

[0049] S4 identifies accelerated events based on the accelerated burst impact index and the accelerated sustained kinetic energy factor.

[0050] Specifically, after determining the acceleration burst index and the acceleration sustained kinetic energy factor, acceleration events can be identified based on these two factors. For example, when both the acceleration burst impact index and the acceleration sustained kinetic energy factor simultaneously exceed a certain threshold, it is determined to be a valid rapid acceleration event; if only the acceleration burst impact index is high while the acceleration sustained kinetic energy factor is low, it is considered an invalid trigger, possibly caused by accidental acceleration or road disturbance; if neither reaches the threshold, it does not constitute an acceleration event. By setting a dual-threshold linkage criterion, the identification results are ensured to simultaneously satisfy the suddenness of the driver's intention and the continuity of the vehicle's power response, thereby achieving reliable capture of real acceleration behavior in complex driving scenarios.

[0051] As a result, the vehicle's longitudinal kinematic parameters, driver operation parameters, and engine power output parameters are all integrated into the same recognition chain. Acceleration recognition no longer relies on a single acceleration threshold, and continuous recognition no longer deviates from the dynamic mechanism. Ultimately, it can distinguish between real rapid acceleration and scenarios such as natural acceleration on slopes and inertial coasting, and output continuous and complete event results.

[0052] According to one embodiment of this application, determining the acceleration burst impact index based on longitudinal acceleration and throttle opening includes: determining a target longitudinal acceleration based on longitudinal acceleration; determining jerk of the target longitudinal acceleration based on the target longitudinal acceleration, wherein jerk represents the instantaneous rate of change of the target longitudinal acceleration; determining the instantaneous rate of change of the throttle opening based on the throttle opening; and determining the acceleration burst impact index based on the jerk and the instantaneous rate of change of the throttle opening.

[0053] Specifically, in order to further optimize the locking accuracy at the start of the acceleration event, when determining the acceleration burst impact index based on longitudinal acceleration and throttle opening, the target longitudinal acceleration is first determined based on longitudinal acceleration; then the jerk of the target longitudinal acceleration is determined based on the target longitudinal acceleration; next, the instantaneous rate of change of throttle opening is determined based on throttle opening; and finally, the acceleration burst impact index is determined based on the jerk and the instantaneous rate of change of throttle opening.

[0054] In this preferred embodiment, the target longitudinal acceleration can be understood as the actual driver-intended acceleration used for subsequent index calculations. After obtaining the target longitudinal acceleration, the jerk can be obtained through methods such as differencing adjacent sampling points, filtering difference, and local fitting differentiation. The jerk represents the instantaneous rate of change of the target longitudinal acceleration and is used to capture acceleration transitions within a short period of time. Simultaneously, the instantaneous rate of change of the throttle opening can be obtained by performing a first-order difference on the throttle opening to characterize the degree of abrupt changes in driver operation. Subsequently, when the jerk pulse and the abrupt change in the throttle rate of change are synchronous within the same time period, the acceleration burst impact index can be determined.

[0055] This constraint bases the initiation determination on both driver operation and vehicle response. The principle is that true rapid acceleration typically begins with the driver rapidly increasing the throttle, causing a sudden increase in driving force, ultimately manifesting as a sharp change in longitudinal acceleration. If only the magnitude of acceleration is observed, it may be difficult to distinguish between gradual and explosive acceleration; if only throttle changes are observed, inconsistencies may arise in scenarios such as revving the engine in neutral or changes in load. Using a combined approach improves the consistency between the initiation determination and the actual intent to accelerate rapidly.

[0056] Those skilled in the art will understand that alternative methods to achieve the same purpose include: first smoothing and filtering the target longitudinal acceleration and then differentiating it, or using a multi-point central difference method to calculate the jerk, or replacing the instantaneous rate of change of throttle opening with the rate of change of throttle control quantity, the rate of change of drive request, and other equivalent quantities.

[0057] For example, the processor can calculate jerk_raw(t) = d(a_true) / dt on a uniform time series according to the sampling interval Δt, where jerk_raw(t) is the jerk of the target's longitudinal acceleration at time t, and a_true is the target's longitudinal acceleration. The instantaneous rate of change of throttle opening can be calculated as [throttle_pos(t) - throttle_pos(t - Δt)] / Δt, where throttle_pos(t) is the throttle opening at time t. These parameters are preferably obtained on a sequence resampled at 100Hz to keep the determination error at the start time within a small range, for example, within approximately 50ms. The relevant parameters are not limited to these and can also be adjusted according to the platform's response speed, sampling accuracy, and controller broadcast cycle.

[0058] Therefore, by adopting the above method, we can more accurately distinguish between sudden acceleration and steady acceleration, which helps to solve the problem of difficulty in locking the speed at the beginning and provides a more accurate starting boundary for subsequent continuous state judgment.

[0059] According to one embodiment of this application, determining the acceleration burst impact index based on the instantaneous rate of change of jerk and throttle opening includes: calculating the effective value of jerk within a first preset time period, and calculating the average value of the instantaneous rate of change of throttle opening within the first preset time period, wherein the effective value is the square root of the average of the squares of each sampled value of jerk; when the effective value is greater than a preset jerk threshold and the average value is greater than a preset throttle abrupt change rate threshold, determining the acceleration burst impact index to characterize the start of a rapid acceleration event. The first preset time period, preset jerk threshold, preset throttle abrupt change rate threshold, and target threshold can be determined according to actual conditions.

[0060] Specifically, to further improve the anti-interference capability of the initial determination, in this preferred embodiment, when determining the acceleration burst impact index based on the instantaneous change rate of acceleration and throttle opening, the effective value of acceleration can be calculated within a first preset time period, and the average value of the instantaneous change rate of throttle opening within the first preset time period can be calculated. When the effective value is greater than a preset acceleration threshold and the average value is greater than a preset throttle change rate threshold, it indicates that the vehicle is in a significant acceleration dynamic process, and this process has sudden and high-intensity characteristics, which conforms to the typical performance of a rapid acceleration event. At this time, the acceleration burst impact index can be initialized to 1, and this acceleration burst impact index represents the start of the rapid acceleration event.

[0061] It should be noted that, in addition to using a Boolean value (e.g., 1 for event triggering, 0 for non-triggering) to represent the start of the aforementioned accelerated event, a non-Boolean initial identifier can also be used, such as a specific numerical value or status code, as long as the identifier can clearly distinguish between the two states of "event not occurring" and "event triggered" in the system logic. This design can provide a more flexible baseline starting point for the continuous evolution of the subsequent accelerated burst impact index.

[0062] In this scheme, the effective value of jerk can be represented by the root mean square (RMS) value, which is calculated by averaging the squares of all sampled values ​​within the window and then taking the square root. Because road bumps, mechanical vibrations, and sampling noise can generate isolated spikes, relying solely on a single instantaneous value can easily lead to false triggering. Quantifying the impact energy using the in-window RMS (Root Mean Square) method preserves the pulse characteristics that truly last for a certain period. Simultaneously, the instantaneous rate of change of throttle opening is represented by the window average, which characterizes the driver's overall operational intensity over that time period, rather than relying on a single outlier.

[0063] For example, the processor can calculate the effective jerk_rms(t) = sqrt(mean(jerk_raw²)) within a sliding window of length t_window_onset (a first preset time), and calculate the average throttle change rate delta_thr_avg(t) over the first preset time. In some embodiments, it can also be required that at least one sampling point within the window satisfies jerk_raw greater than a peak threshold, thereby balancing peak characteristics and window statistical stability. As an example, t_window_onset can be 0.2s, the preset jerk threshold can correspond to a calibration value around 3.25 m / s³, and the preset throttle change rate threshold can correspond to a calibration value around 40% / s. The above values ​​are used to illustrate an optional calibration method, and in practical applications, they can be adjusted within a certain range according to the characteristics of the vehicle platform.

[0064] Those skilled in the art will understand that the first preset time is not limited to 0.2s, and can be set to other values ​​between 0.1s and 0.5s depending on the sampling frequency, vehicle dynamic response, and vibration level; the effective value is not limited to the root mean square value, and can also be the absolute value average, energy integral, or other statistics that can reflect the degree of impact within the window; the average value of the instantaneous rate of change of throttle opening can also be replaced by the median, quantile, or weighted average. These variations can all be used to suppress false triggering caused by high-frequency interference.

[0065] Therefore, by adopting the above-mentioned window statistics and dual threshold judgment method, high-frequency disturbances caused by road noise and vehicle vibration can be smoothed, the consistency and stability of the initial judgment can be improved, and the probability of false triggering caused by isolated spikes can be reduced.

[0066] According to one embodiment of this application, the engine power output parameters include engine torque and engine speed. Determining the acceleration sustained kinetic energy factor based on longitudinal acceleration, engine power output parameters, and vehicle mass includes: determining a target longitudinal acceleration based on the longitudinal acceleration; determining engine output power based on the product of engine torque and engine speed; determining specific power based on the ratio of engine output power to vehicle mass; and weighted summing the ratio of the target longitudinal acceleration to a preset acceleration reference threshold and the ratio of specific power to a preset specific power reference threshold to obtain the acceleration sustained kinetic energy factor.

[0067] Specifically, to further enhance the verification of the authenticity of the continuous active acceleration process, in this preferred embodiment, the engine power output parameters may include engine torque and engine speed. By combining these two parameters to calculate the engine output power, and further correlating it with the vehicle mass, the power input level obtained by the vehicle per unit mass is reflected. This output power works in conjunction with longitudinal acceleration to construct the continuous acceleration kinetic energy factor, thereby distinguishing between brief disturbances and real, continuous active acceleration behavior.

[0068] When determining the acceleration sustained kinetic energy factor based on longitudinal acceleration, engine power output parameters, and vehicle mass, the target longitudinal acceleration is first determined based on the longitudinal acceleration. Then, the engine output power is determined based on the product of engine torque and engine speed. Next, the specific power is determined based on the ratio of engine output power to vehicle mass. Finally, the ratio of the target longitudinal acceleration to the preset acceleration reference threshold and the ratio of the specific power to the preset specific power reference threshold are weighted and summed to obtain the acceleration sustained kinetic energy factor.

[0069] In this preferred embodiment, engine output power can be calculated from torque and speed, while specific power further eliminates the bias when directly comparing vehicles of different masses. Target longitudinal acceleration characterizes the vehicle's current perceived longitudinal strength, while specific power characterizes the power source's ability to support this acceleration state. Normalizing and fusing these two parameters allows the resulting factor to simultaneously reflect both kinematic results and dynamic causes. When the vehicle is on a downhill section, even with high longitudinal acceleration, if the engine's torque and speed-corresponding output power are insufficient, the specific power component contributes less, and the overall factor is less likely to maintain its condition. Conversely, when the driver actively applies heavy throttle and the engine continuously outputs high power, the factor is more likely to remain at a high level.

[0070] For example, the processor can calculate sp(t) = (torque_act(t) × rpm_eng(t) / 9549) / mass_est, and use a_base = 0.8 m / s² and sp_base = 50 W / kg as reference benchmarks to normalize the target longitudinal acceleration and specific power. Then, the sustained kinetic energy factor is obtained through haki(t) = α·(a_true(t) / a_base) + β·(sp(t) / sp_base), where α and β can satisfy α + β = 1, preferably α = 0.6 and β = 0.4. These parameters can be recalibrated according to the vehicle's power level, weight range, and riding characteristics.

[0071] In addition, alternative methods to achieve the same purpose include: using wheel-end power instead of engine-side power, using real-time estimated vehicle mass instead of fixed mass value, or applying amplitude limiting, filtering, or nonlinear compression processing to acceleration and specific power before weighted summation. As long as a joint factor that takes into account both the feeling of being pushed back into the seat and the power output of the power source can be formed, it can be used to characterize continuous power output capability.

[0072] Therefore, active acceleration and inertial slope can be fundamentally distinguished, thereby improving the accuracy of in-process state determination.

[0073] According to one embodiment of this application, determining a target longitudinal acceleration based on longitudinal acceleration includes: obtaining the longitudinal slope angle of the current road; and using the longitudinal slope angle to perform slope compensation on the longitudinal acceleration to obtain the target longitudinal acceleration.

[0074] Specifically, in order to reduce the interference of the slope on subsequent calculations, in this preferred scheme, when determining the target longitudinal acceleration based on the longitudinal acceleration, the longitudinal slope angle of the current road can be obtained, and the longitudinal slope angle can be used to perform slope compensation on the longitudinal acceleration to obtain the target longitudinal acceleration.

[0075] The base longitudinal slope angle can be obtained from attitude calculations by a six-axis inertial measurement unit (IMU), or it can be provided by map slope information, vehicle attitude estimation results, or multi-sensor fusion results. The raw longitudinal acceleration usually includes the component of gravity in the vehicle's longitudinal direction; therefore, in uphill and downhill scenarios, directly using the raw value can easily deviate from the driver's true acceleration intention. By performing slope compensation, the gravity component caused by road inclination can be eliminated, making the target longitudinal acceleration closer to the equivalent active acceleration of the vehicle on a straight road.

[0076] For example, the processor can calculate the compensated target longitudinal acceleration based on a_true = a_long - g × sin(slope_ang), where g is the gravitational acceleration constant and slope_ang is the longitudinal slope angle of the current road. In some embodiments, slope_ang can be obtained by filtering the attitude information output by the IMU. After the above compensation, subsequent jerk, continuous kinetic energy factor, and state judgment can be based on inputs that are closer to the driver's actual intention.

[0077] Those skilled in the art will understand that the method of obtaining the slope angle is not limited to a single IMU calculation; it can also incorporate satellite elevation changes, map slope priors, wheel speed and attitude fusion estimation, etc.; the compensation formula can also be modified by combining installation offset and attitude drift compensation coefficients. These modified methods are all equivalent implementations of removing the gravity component from the original longitudinal acceleration.

[0078] This significantly reduces the false alarm and false negative rates caused by uphill and downhill sections, and provides a unified and reliable target longitudinal acceleration input for starting, continuing, and ending identification. Furthermore, the slope compensation scheme in this embodiment can be combined with the aforementioned window statistical judgment and weighted fusion judgment schemes to achieve a more stable overall identification effect.

[0079] According to one embodiment of this application, identifying acceleration events based on the acceleration burst impact index and the acceleration sustained kinetic energy factor includes: when the acceleration burst impact index is set to indicate the start of an acceleration event, determining that the vehicle has started to accelerate and recording the acceleration time; when the vehicle has started to accelerate, if the acceleration sustained kinetic energy factor is greater than a first preset threshold and the target longitudinal acceleration is greater than a preset minimum acceleration threshold, then determining that the vehicle is in an acceleration state and recording the corresponding acceleration parameters.

[0080] Specifically, to facilitate the transition of an event from its initial triggering to a sustained output state management process, in this preferred embodiment, when identifying acceleration events based on the acceleration burst impact index and the acceleration sustained kinetic energy factor, the acceleration burst index is judged. If the acceleration burst impact index indicates the start of a rapid acceleration event, it can be determined that the vehicle has begun acceleration, and the current acceleration moment is recorded. Once the vehicle has begun acceleration, the magnitudes of the acceleration sustained kinetic energy factor and the target longitudinal acceleration are judged. If the acceleration sustained kinetic energy factor is greater than a first preset threshold and the target longitudinal acceleration is greater than a preset minimum acceleration threshold, then the vehicle is determined to be in an acceleration state, and the corresponding acceleration parameters are recorded.

[0081] The acceleration burst impact index is used to trigger the start, while the acceleration sustained kinetic energy factor and the target longitudinal acceleration are used together to maintain the ongoing state. This design is because once an event is triggered by the start signal, there may still be brief false activations. Therefore, the sustained kinetic energy factor and the minimum acceleration threshold are needed to further confirm whether the event has entered a true sustained phase. In other words, the start signal focuses more on burst detection at the initial moment, while the ongoing determination focuses more on the authenticity and continuity of the state.

[0082] For example, when the acceleration burst impact index haoi is set to the target threshold, such as jumping from 0 to 1, the processor can record the acceleration time t_start and switch the state machine from the idle state to the active state. After entering the active state, if the acceleration sustained kinetic energy factor haki is greater than the first preset threshold and the target longitudinal acceleration a_true is greater than the preset minimum acceleration threshold, then it can be determined that the vehicle is in an acceleration state, and acceleration parameters are continuously recorded. The acceleration parameters here may include peak acceleration, maximum specific power, maximum sustained kinetic energy factor, event duration, cumulative power integral, etc. In some embodiments, the first preset threshold may be around 1.2, and the preset minimum acceleration threshold may be around 0.6 m / s².

[0083] Those skilled in the art will understand that the acceleration parameters are not limited to those listed above, and may also include average acceleration, average throttle opening, peak speed, peak torque, integral mileage, or information related to vehicle identification; the first preset threshold and the minimum acceleration threshold may also be adaptively adjusted according to vehicle model, displacement, or load. These variations do not affect the basic idea of ​​outputting a complete event through start triggering and continuous confirmation.

[0084] Therefore, by adopting the above scheme, we can avoid only outputting scattered trigger points, but instead organize the start to the continuous phase into manageable event objects, thereby making the recognition results more complete in both the time and parameter dimensions.

[0085] According to one embodiment of this application, the vehicle acceleration event recognition method further includes: recording the duration when the acceleration duration kinetic energy factor is less than a first preset threshold and greater than a second preset threshold; and maintaining the vehicle in an acceleration state if the duration is less than or equal to a preset duration threshold. The preset duration threshold can be determined according to actual conditions.

[0086] Specifically, in order to adapt to the short-term power interruption during vehicle gear shifting, in this preferred embodiment, when the acceleration continuous kinetic energy factor is less than a first preset threshold and greater than a second preset threshold, the duration is recorded; if the duration is less than or equal to the preset duration threshold, the vehicle is kept in an acceleration state.

[0087] In this scheme, the area above the first preset threshold is considered the normal maintenance range, the area below the second preset threshold is considered the obvious exit trend range, and the intermediate range between the two is considered the tolerance range. During continuous rapid acceleration, gear shifting may cause a power interruption of approximately 200ms to 400ms, during which longitudinal acceleration and continuous kinetic energy factor may momentarily drop. If the vehicle exits the state immediately at this moment, a continuous acceleration may be fragmented into multiple fragmented events. By setting a duration protection within the intermediate range, the gear shift gap can be tolerated.

[0088] For example, when the processor is active, it can monitor whether the acceleration sustained kinetic energy factor (haki) drops from a normal maintenance range greater than a first preset threshold to a tolerance range between the first and second preset thresholds. For example, when haki is briefly between 0.8 and 1.2, a duration timer can be started; if the duration does not exceed a preset duration threshold, such as 0.3 seconds, the state machine remains active and the event is not terminated. If the threshold is exceeded, other exit conditions can be combined to determine whether to exit. The above parameters can be preferably adapted to the short-term power drop characteristics caused by gear shifting in a vehicle's sequential gearbox.

[0089] Those skilled in the art will understand that the second preset threshold and the preset duration threshold are not limited to the example values ​​mentioned above, and can be adjusted according to the powertrain response, shift strategy, and sampling resolution; the intermediate tolerance range can also be designed as a multi-level range, or an adaptive range that varies with vehicle speed, gear, or engine speed. Any approach that can maintain state continuity in the event of a short-term power drop but before the event truly ends is applicable.

[0090] Therefore, by adopting the above scheme, erroneous exits caused by gear shift gaps can be avoided, thereby reducing event fragmentation and improving the consistency of statistical standards.

[0091] According to one embodiment of this application, after determining that the vehicle is in an acceleration state, the vehicle acceleration event recognition method further includes: acquiring the vehicle speed, and calculating the average value of the acceleration jerk of the target longitudinal acceleration within a second preset time period; determining that the vehicle exits the acceleration state based on the average value, throttle opening, and vehicle speed. The second preset time period can be determined according to actual conditions.

[0092] Specifically, in order to determine the exit time more promptly before and after the acceleration sensation disappears, in this preferred embodiment, after determining that the vehicle is in an acceleration state, the vehicle speed can also be obtained, and the average value of the acceleration jerk of the target longitudinal acceleration can be calculated within a second preset time. Then, based on the average value, throttle opening and vehicle speed, the vehicle exits the acceleration state.

[0093] In this scheme, the exit judgment no longer relies solely on the decrease in acceleration value, but combines the acceleration trend, throttle release, and the stabilization of vehicle speed. The principle is that the end of rapid acceleration is usually a transition process of releasing the throttle, decelerating, and reaching a steady speed. If the end is only determined when the acceleration falls below a fixed threshold, there may be a lag in judgment when engine braking generates negative acceleration. By observing whether the average acceleration value turns negative within a second preset time period, the inflection point of the acceleration slope changing from positive to negative can be identified; combined with throttle opening and vehicle speed stability, it is possible to determine earlier that the feeling of acceleration has begun to disappear.

[0094] For example, the processor can acquire the vehicle speed v and calculate the average jerk of the target longitudinal acceleration within a time window of length t_window_fade (a second preset time), as the average slope of the acceleration change slope, slope_a(t), which can be equivalent to the average value of jerk_raw within the window. If this average value is less than a preset jerk threshold, and the throttle opening has been reduced to a small level, and the vehicle speed change tends to converge, then an exit candidate signal can be generated. In some embodiments, t_window_fade can be taken around 0.4s to balance sensitivity to trend changes and suppression of short-term spikes.

[0095] Those skilled in the art will understand that the second preset time can be adjusted according to the vehicle's power response and filtering requirements; the vehicle speed used to characterize the speed state can come from the satellite positioning module, the wheel speed interface, or a combination of both; the average jerk can also be obtained using weighted average, sliding fit slope, or other equivalent forms. As long as trend information, throttle state, and vehicle speed state can be used together to determine exit, a similar effect can be achieved.

[0096] Therefore, by adopting the above scheme, the exit prediction can be completed more promptly before and after the feeling of rapid acceleration truly disappears. In some embodiments, the termination trend can be identified about 200ms earlier than the simple acceleration threshold method, thus making the exit smoother and more stable.

[0097] According to one embodiment of this application, determining whether a vehicle is exiting acceleration mode based on an average value, throttle opening, and vehicle speed includes: determining the vehicle speed fluctuation rate based on the vehicle speed, where the vehicle speed fluctuation rate represents the instantaneous change in vehicle speed as it approaches steady-state cruising; and determining that the vehicle is exiting acceleration mode when the average value is less than a preset acceleration threshold, the throttle opening is less than a preset cruising throttle opening threshold, the vehicle speed fluctuation rate is less than a preset vehicle speed fluctuation rate threshold, and this condition persists for a preset duration threshold. The preset acceleration threshold, preset cruising throttle opening threshold, preset vehicle speed fluctuation rate threshold, and preset duration threshold can be determined according to actual conditions.

[0098] Specifically, in order to improve the reliability of the exit determination, in this preferred embodiment, when determining the vehicle exits the acceleration state based on the average value, throttle opening and vehicle speed, the vehicle speed fluctuation rate is first determined based on the vehicle speed. When the average value is less than the preset acceleration threshold, the throttle opening is less than the preset cruise throttle opening threshold, the vehicle speed fluctuation rate is less than the preset vehicle speed fluctuation rate threshold and continues for a preset duration threshold, the vehicle exits the acceleration state.

[0099] In this scheme, vehicle speed volatility can be used to characterize whether the vehicle speed tends towards steady-state cruising. Even if the negative trend, low throttle, and low volatility conditions are met only at a single moment, the problem may still be caused by occasional disturbances or road bumps. By requiring multiple conditions to be met continuously over a period of time, the stability of the exit confirmation can be significantly improved. This sustained holding mechanism is equivalent to adding a de-jittering process to the exit candidate signal, making the final exit time closer to the actual moment when the throttle is released and steady-state cruising begins.

[0100] It should be noted that steady-state cruise control refers to the state in which a vehicle travels along a road at a relatively constant speed without any obvious intention to accelerate or decelerate. In this state, the throttle opening remains basically stable, the output power of the engine or drive motor fluctuates little, the rate of change of vehicle speed (i.e., acceleration) approaches zero, and the rate of speed fluctuation is below a preset threshold. This state typically occurs when the driver releases the accelerator pedal after accelerating and the vehicle maintains its speed relying on inertia or the cruise control system. It is a key reference point for determining whether a rapid acceleration event has ended.

[0101] For example, the processor can calculate the vehicle speed fluctuation rate based on the vehicle speed sequence, such as using |dv / dt| to represent the degree of instantaneous change. When the average acceleration within the window is less than the preset acceleration threshold thr_jerk_negative, the throttle opening is less than the preset cruise throttle opening threshold thr_throttle_cruise, and the vehicle speed fluctuation rate is less than the preset vehicle speed fluctuation rate threshold thr_v_fluctuation, and the above conditions are continuously maintained for a preset duration threshold t_hold_exit, the vehicle can be considered to have exited the acceleration state. In some embodiments, thr_jerk_negative can be around -0.5 m / s³, thr_throttle_cruise can be around 8%, thr_v_fluctuation can be around 0.3 m / s², and t_hold_exit can be around 0.3 s. To reduce the offset of the anti-shake delay from the end boundary, the end time can also be recorded as t_end = t - t_hold_exit, so that the result is closer to the actual throttle release point.

[0102] Those skilled in the art will understand that vehicle speed fluctuation rate is not limited to the absolute value of the first derivative, but can also be expressed as speed standard deviation, windowed variance, frequency domain fluctuation, etc.; the upper limit of cruise throttle opening and the threshold of vehicle speed fluctuation rate can also be adjusted according to vehicle type, driving mode, and load status. The selection of continuous hold time can be a trade-off between suppressing glitches and ensuring response speed.

[0103] Therefore, by adopting the above exit confirmation scheme, we can avoid erroneous exits caused by momentary disturbances, make the end time more stable, and further improve the availability of complete event output.

[0104] In summary, as a concrete example, the vehicle's onboard terminal, through a six-axis inertial measurement unit, CAN bus interface, satellite positioning module, and wheel speed sensor interface, acquires real-time longitudinal acceleration, throttle opening, engine torque, engine speed, vehicle speed, and estimated vehicle mass. Data from different sampling frequencies are then resampled to 100 Hz using a unified timestamp. During vehicle operation, the terminal processor executes a program that continuously outputs complete rapid acceleration events.

[0105] Taking starting at a traffic light in an urban area as an example. The driver quickly twists the accelerator lever. The initial longitudinal acceleration is first compensated for by the longitudinal slope angle. The processor uses a six-axis inertial measurement unit to calculate the slope angle, eliminating the gravity component of the slope by multiplying the target longitudinal acceleration by the sine of the slope angle. Subsequently, the processor calculates the jerk based on the target longitudinal acceleration and simultaneously calculates the instantaneous rate of change of the throttle opening. Within a 0.2-second sliding window, if the effective value of the jerk exceeds 3.25 m / s³ and the average rate of change of the throttle within the window exceeds 40% / second, the acceleration burst impact index jumps from 0 to 1, the state machine switches from idle to active state, and the acceleration start time is recorded. Due to the use of window statistics and dual threshold judgment, high-frequency road vibrations and isolated spikes will not directly trigger the start, and the start time error can be controlled within approximately 50 milliseconds.

[0106] Once activated, the processor calculates the engine output power using engine torque and engine speed, and divides it by the vehicle's mass to obtain the specific power. The target longitudinal acceleration is compared to a baseline acceleration value of 0.8 m / s², and the specific power is compared to a baseline specific power value of 50 watts / kg. These ratios are then weighted and summed with weights of 0.6 and 0.4 respectively to obtain the acceleration sustaining kinetic energy factor. When this factor is greater than 1.2 and the target longitudinal acceleration is greater than 0.6 m / s², the vehicle remains in an acceleration state, and the processor continuously updates the peak acceleration, maximum specific power, and cumulative power integral. If a gear shift occurs mid-acceleration, the sequential gearbox causes a power interruption of approximately 200 to 400 milliseconds, and this factor may briefly drop back to between 0.8 and 1.2. In this case, the processor records the duration; as long as it does not exceed 0.3 seconds, the state machine remains active and does not exit the event, thus preventing a continuous acceleration from being fragmented into multiple parts.

[0107] When the driver releases the accelerator, the processor calculates the average longitudinal acceleration within a 0.4-second trend judgment window, while simultaneously monitoring the throttle opening and the vehicle speed fluctuation rate calculated based on vehicle speed. If the average value is less than -0.5 m / s³, the throttle opening is less than 8%, and the vehicle speed fluctuation rate is less than 0.3 m / s², and these conditions are maintained continuously for 0.3 seconds, the vehicle exits the acceleration state. The state machine switches to the end state, records the end time, and compensates for anti-shake delay to make the end time closer to the actual throttle release point. The final output event log includes parameters such as start time, end time, duration, peak acceleration, maximum specific power, and maximum sustained kinetic energy factor, and is pushed to the cloud queue.

[0108] In this scenario, the accuracy of driving behavior scoring and the statistical caliber of continuous events are significantly improved. For example, on long downhill or coasting sections, even if the vehicle achieves high longitudinal acceleration, as long as the engine torque and throttle opening do not respond synchronously, the continuous kinetic energy factor is difficult to maintain, thus avoiding false alarms of rapid acceleration. Conversely, when accelerating sharply uphill, the target longitudinal acceleration after gradient compensation can prevent underreporting. For insurance pricing and rider safety management, the platform no longer receives scattered threshold exceedances, but rather complete events with true start and end boundaries and process extreme values, which can more fairly reflect the degree of aggressive driving by the driver.

[0109] The following is combined Figure 2 The method described in this application is used to describe the method.

[0110] As a specific example, the vehicle acceleration event recognition method of this application may include the following steps: S101, obtain the vehicle's longitudinal acceleration, throttle opening, engine torque, engine speed and vehicle mass, and proceed to steps S102 and S106 respectively.

[0111] S102, determine the target longitudinal acceleration based on the longitudinal acceleration, determine the jerk of the target longitudinal acceleration based on the target longitudinal acceleration, and determine the instantaneous rate of change of the throttle opening based on the throttle opening.

[0112] S103, calculate the effective value of the acceleration within the first preset time period, and calculate the average value of the instantaneous rate of change of the throttle opening within the first preset time period.

[0113] S104. When the effective value is greater than the preset jerk threshold and the average value is greater than the preset throttle change rate threshold, the acceleration burst impact index is determined to represent the start of the rapid acceleration event.

[0114] S105, confirm that the vehicle has started to accelerate, record the acceleration time, and proceed to step S108.

[0115] S106 determines the target longitudinal acceleration based on the longitudinal acceleration and determines the engine output power based on the product of engine torque and engine speed.

[0116] S107. The specific power is determined based on the ratio of engine output power to vehicle mass. The ratio of target longitudinal acceleration to preset acceleration reference threshold and the ratio of specific power to preset specific power reference threshold are weighted and summed to obtain the acceleration continuous kinetic energy factor.

[0117] S108, determine whether the acceleration sustained kinetic energy factor is greater than the first preset threshold and whether the target longitudinal acceleration is greater than the preset minimum acceleration threshold. If yes, proceed to step S109; if no, proceed to step S112.

[0118] S109, Determine that the vehicle is in an acceleration state and record the corresponding acceleration parameters.

[0119] S110: Obtain the vehicle speed, calculate the average value of the acceleration of the target longitudinal acceleration within a second preset time period, and determine the vehicle speed fluctuation rate based on the vehicle speed.

[0120] S111, when the average value is less than the preset acceleration threshold, the throttle opening is less than the preset cruise throttle opening threshold, the vehicle speed fluctuation rate is less than the preset vehicle speed fluctuation rate threshold, and the vehicle speed continues to be less than the preset duration threshold, the vehicle is determined to exit the acceleration state.

[0121] S112, determine whether the acceleration continuous kinetic energy factor is less than the first preset threshold and greater than the second preset threshold. If yes, proceed to step S113; if no, proceed to step S101.

[0122] S113, record the duration, and if the duration is less than or equal to a preset duration threshold, maintain the vehicle in an acceleration state and proceed to step S110.

[0123] In summary, the vehicle acceleration event recognition method according to the embodiments of this application acquires the vehicle's longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass. Based on the longitudinal acceleration and throttle opening, an acceleration burst impact index is determined, representing the intensity of the driver's intent to accelerate rapidly. Based on the longitudinal acceleration, engine power output parameters, and vehicle mass, an acceleration sustaining kinetic energy factor is determined, representing the strength of the vehicle's sustained power output capability. Acceleration events are then identified based on the acceleration burst impact index and the acceleration sustaining kinetic energy factor. Therefore, this method can reduce false alarms and false negatives of non-real rapid acceleration, achieving accurate and continuous recognition of real rapid acceleration events.

[0124] Corresponding to the above embodiments, this application also proposes a vehicle.

[0125] like Figure 3 As shown, the vehicle 200 in this embodiment may include: a memory 210, a processor 220, and a program stored in the memory 210 and executable on the processor 220. When the processor 220 executes the program, it implements the above-described vehicle acceleration event recognition method.

[0126] The vehicle according to the embodiments of this application, by executing the above-described vehicle acceleration event identification method, can reduce false alarms and false negatives of non-real rapid acceleration, and achieve accurate and continuous identification of real rapid acceleration events.

[0127] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0128] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0129] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0130] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0131] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for recognizing vehicle acceleration events, characterized in that, The method includes: The longitudinal acceleration, throttle opening, engine power output parameters, and vehicle mass of the vehicle are obtained. The acceleration burst impact index is determined based on the longitudinal acceleration and the throttle opening, and the acceleration burst impact index represents the burst intensity of the driver's intention to accelerate rapidly; The acceleration sustained kinetic energy factor is determined based on the longitudinal acceleration, the engine power output parameters, and the vehicle mass. The acceleration sustained kinetic energy factor represents the strength of the vehicle's sustained power output capability. The acceleration event is identified based on the acceleration burst impact index and the acceleration sustained kinetic energy factor.

2. The vehicle acceleration event recognition method according to claim 1, characterized in that, The determination of the acceleration burst impact index based on the longitudinal acceleration and the throttle opening includes: Determine the target longitudinal acceleration based on the aforementioned longitudinal acceleration; The jerk of the target longitudinal acceleration is determined based on the target longitudinal acceleration, and the jerk represents the instantaneous rate of change of the target longitudinal acceleration; The instantaneous rate of change of the throttle opening is determined based on the throttle opening. The acceleration burst impact index is determined based on the instantaneous rate of change of the jerk and the throttle opening.

3. The vehicle acceleration event recognition method according to claim 2, characterized in that, The determination of the acceleration burst impact index based on the instantaneous rate of change of the jerk and the throttle opening includes: The effective value of the acceleration is calculated within a first preset time period, and the average value of the instantaneous rate of change of the throttle opening is calculated within the first preset time period. The effective value is the square root of the average of the squares of the various sampled values ​​of the acceleration. When the effective value is greater than a preset jerk threshold and the average value is greater than a preset throttle change rate threshold, the acceleration burst impact index is determined to represent the start of the rapid acceleration event.

4. The vehicle acceleration event recognition method according to claim 1, characterized in that, The engine power output parameters include engine torque and engine speed. The determination of the acceleration sustained kinetic energy factor based on the longitudinal acceleration, the engine power output parameters, and the vehicle mass includes: Determine the target longitudinal acceleration based on the aforementioned longitudinal acceleration; The engine output power is determined based on the product of the engine torque and the engine speed; The specific power is determined based on the ratio of the engine output power to the vehicle mass; The acceleration sustained kinetic energy factor is obtained by weighted summing the ratio of the target longitudinal acceleration to a preset acceleration reference threshold and the ratio of the specific power to a preset specific power reference threshold.

5. The vehicle acceleration event recognition method according to claim 2 or 4, characterized in that, Determining the target longitudinal acceleration based on the longitudinal acceleration includes: Obtain the longitudinal slope angle of the current road; The longitudinal acceleration is obtained by using the longitudinal slope angle to perform slope compensation on the longitudinal acceleration.

6. The vehicle acceleration event recognition method according to claim 1, characterized in that, The identification of the acceleration event based on the acceleration burst impact index and the acceleration sustained kinetic energy factor includes: When the acceleration burst impact index characterizes the start of the rapid acceleration event, it is determined that the vehicle begins to accelerate, and the acceleration moment is recorded. When the vehicle begins to accelerate, if the acceleration sustained kinetic energy factor is greater than a first preset threshold and the target longitudinal acceleration is greater than a preset minimum acceleration threshold, then the vehicle is determined to be in an acceleration state, and the corresponding acceleration parameters are recorded.

7. The vehicle acceleration event recognition method according to claim 6, characterized in that, The method further includes: If the acceleration sustained kinetic energy factor is less than the first preset threshold and greater than the second preset threshold, the duration is recorded. If the duration is less than or equal to a preset duration threshold, the vehicle remains in an acceleration state.

8. The vehicle acceleration event recognition method according to claim 6, characterized in that, After determining that the vehicle is in an accelerating state, the method further includes: The vehicle speed is obtained, and the average value of the longitudinal acceleration of the target is calculated within a second preset time period; The vehicle exits the acceleration state based on the average value, the throttle opening, and the vehicle speed.

9. The vehicle acceleration event recognition method according to claim 8, characterized in that, The step of determining whether the vehicle exits the acceleration state based on the average value, the throttle opening, and the vehicle speed includes: The vehicle speed fluctuation rate is determined based on the vehicle speed, and the vehicle speed fluctuation rate represents the degree of instantaneous change of the vehicle speed when it tends towards steady-state cruising. When the average value is less than a preset acceleration threshold, the throttle opening is less than a preset cruise throttle opening threshold, and the vehicle speed fluctuation rate is less than a preset vehicle speed fluctuation rate threshold and continues for a preset duration threshold, the vehicle is determined to exit the acceleration state.

10. A vehicle, characterized in that, include: The device includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the vehicle acceleration event recognition method according to any one of claims 1-9.

Images