An adaptive correction method and system for a horse gait cycle

Abstract

This invention belongs to the field of equine movement monitoring technology, specifically relating to an adaptive correction method and system for equine gait cycles. The method first uses sensors to collect triaxial acceleration signals, angular velocity signals, and magnetic field vectors in real time during horse movement, and uses the magnetic field vector to correct the sensor's spatial coordinate system. It then collects the horse's movement speed and speed gradient in real time, detecting whether the speed gradient triggers the gait reorganization threshold. If not triggered, a step counting time window is calculated based on the horse's movement speed, and steps are counted according to the acceleration threshold and the step counting time window. If triggered, the acceleration threshold is linearly lowered based on the speed gradient, and the step counting time window is calculated again based on the horse's movement speed. Steps are counted according to the lowered acceleration threshold and the step counting time window. This invention effectively solves the problem of missed or incorrect counting of gaits such as slow hops and shuffling gaits in traditional fixed-threshold algorithms.

Description

Technical Field

[0001] This invention belongs to the field of horse movement monitoring technology, specifically relating to an adaptive correction method and system for horse gait cycles. Background Technology

[0002] With the increasing intensity of horse racing and professional training, real-time monitoring of a horse's gait has become a core aspect of its movement. Currently, monitoring moving objects primarily employs inertial navigation technology and common step-counting algorithms. However, racehorses possess physical characteristics significantly different from ordinary moving targets under extreme conditions, leading to the following problems in the practical application of existing technologies:

[0003] 1. Signal distortion under high-speed non-stationary motion. Existing step counting algorithms are mostly based on steady-state dynamic models of the human body or ordinary low-speed targets. When racehorses run at extremely high speeds, their gait undergoes a fundamental physiological reorganization (such as asymmetrical hoof landing during galloping), which manifests as multiple characteristic peaks highly overlapping in the time domain in the acceleration signal. Traditional fixed threshold detection or simple mode switching algorithms cannot deconstruct this complex nonlinear characteristic, easily leading to serious undercounting and miscounting, and making it difficult to accurately reproduce the horse's gait frequency parameters under extreme motion.

[0004] 2. Sensor failure due to strong transient overload. During high-speed training, horses experience a huge instantaneous impact force (High-Gshock) when their hooves touch the ground. Existing multi-sensor fusion solutions typically serve general position navigation or attitude measurement, aiming for smooth output of absolute coordinates. However, this navigation-level logic often suffers from severe logic fluctuations due to signal saturation or integration errors when dealing with the continuous pulse-like overloads unique to horses, causing gait feature points to be submerged in noise.

[0005] 3. A disconnect between step counting requirements and navigation model logic. Existing geomagnetic-assisted technologies are mostly defined as geographic location references used to correct spatial drift. However, in horse step counting scenarios, the goal of sensor fusion is not trajectory reconstruction, but the accurate capture of gait phase. When the accelerometer enters a signal blind zone due to a severe impact, existing technology lacks a dedicated correction logic that can utilize the subtle magnetic field phase changes of the horse's torso during long-period movements to serve as physical anchor points for the gait window.

[0006] In conclusion, existing motion measurement technologies, due to their lack of consideration for the specific biomechanical characteristics of horses, fail to meet the requirements of professional competition in terms of both step counting accuracy and logical continuity during high-speed, high-load, and multi-gait transition horse training monitoring. Summary of the Invention

[0007] The purpose of this invention is to address the aforementioned problems in the prior art by providing an adaptive correction method and system for the horse gait cycle with high step counting accuracy in the face of horse gait transitions.

[0008] To achieve the above objectives, the technical solution of the present invention is as follows:

[0009] In a first aspect, the present invention provides an adaptive correction method for horse gait cycles, the adaptive correction method comprising:

[0010] S1. The sensor collects the three-axis acceleration signal, angular velocity signal and magnetic field vector of the horse in real time, and uses the magnetic field vector to correct the spatial coordinate system of the sensor.

[0011] S2. Real-time acquisition of the horse's movement speed and speed gradient, detection of whether the speed gradient triggers the gait reorganization admission condition. If not triggered, proceed to S3; if triggered, proceed to S4.

[0012] S3. Calculate the step counting time window based on the horse's movement speed, and count steps according to the acceleration threshold and the step counting time window.

[0013] S4. Based on the velocity gradient, linearly lower the acceleration judgment threshold and calculate the step counting time window based on the horse's movement speed. Count steps according to the lowered acceleration judgment threshold and the step counting time window.

[0014] The step counting in S3 includes:

[0015] The acceleration magnitude is calculated based on the triaxial acceleration signal during horse movement. It is then determined whether the acceleration magnitude exceeds an acceleration threshold, which is the acceleration threshold corresponding to low-speed steady state. If the acceleration magnitude exceeds the acceleration threshold and the time interval from the previous step pulse meets the step counting time window, it is determined to be a single hoof landing signal. The counter accumulates the hoof landing signal until the step counting cycle is reached. The final expression for the real-time gait frequency is: ,in For real-time cadence, To accumulate the number of hoof landings, This is the step counting cycle.

[0016] The step counting in S4 includes:

[0017] The acceleration magnitude is calculated based on the triaxial acceleration signal of the horse during movement. The primary and secondary peaks of the acceleration magnitude are extracted based on a down-adjusted acceleration threshold. Each primary and secondary peak is identified as a single hoof-drop signal. Whenever a primary or secondary peak is extracted, signal fluctuations below the current primary or secondary peak are filtered out within a step counting window. When the triaxial acceleration signal is lost due to saturation, the horse's rotational envelope is calculated based on the geomagnetic vector, and the rotational envelope is time-aligned with the acceleration magnitude. Step compensation is performed based on the positive peak of the pitch angle within the rotational envelope. The final expression for the real-time gait frequency is: ,in For real-time cadence, The number of times the hoof lands corresponding to the main peak. This represents the number of times the secondary peak is hit. The number of hoof landings for step compensation; This is the step counting period;

[0018] The step counting time window is adjusted based on the following formula, according to the movement speed: ,in Indicates the step counting time window. Indicates average stride length. This indicates the speed at which the horse moves.

[0019] The threshold is determined by linearly reducing the acceleration based on the velocity gradient using the following formula: ,in The threshold for determining the reduced acceleration. The acceleration threshold corresponding to low-speed steady state. This is the horse gait attenuation coefficient. This represents the velocity gradient.

[0020] The gait re-admission condition refers to the velocity gradient exceeding a preset velocity gradient threshold.

[0021] Secondly, the present invention provides an adaptive correction system for horse gait cycles, the adaptive correction system comprising a signal acquisition module, a gait judgment module, a first step counting module, and a second step counting module;

[0022] The signal acquisition module is used to acquire the three-axis acceleration signal, angular velocity signal and magnetic field vector of the horse in real time through the sensor, and to use the magnetic field vector to correct the spatial coordinate system of the sensor.

[0023] The gait judgment module is used to collect the horse's movement speed and speed gradient in real time, and detect whether the speed gradient triggers the gait reorganization admission condition. If it is not triggered, the first step counting module is used to count the steps; if it is triggered, the second step counting module is used to count the steps.

[0024] The first step counting module is used to calculate the step counting time window based on the horse's movement speed, and to count steps according to the acceleration threshold and the step counting time window;

[0025] The second step counting module is used to linearly lower the acceleration judgment threshold based on the speed gradient and calculate the step counting time window based on the horse's movement speed, and count steps according to the lowered acceleration judgment threshold and the step counting time window.

[0026] The first step counting module counts steps according to the following steps:

[0027] The acceleration magnitude is calculated based on the triaxial acceleration signal during horse movement. It is then determined whether the acceleration magnitude exceeds an acceleration threshold, which is the acceleration threshold corresponding to low-speed steady state. If the acceleration magnitude exceeds the acceleration threshold and the time interval from the previous step pulse meets the step counting time window, it is determined to be a single hoof landing signal. The counter accumulates the hoof landing signal until the step counting cycle is reached. The final expression for the real-time gait frequency is: ,in For real-time cadence, To accumulate the number of hoof landings, This is the step counting period;

[0028] The second step counting module counts steps according to the following steps:

[0029] The acceleration magnitude is calculated based on the triaxial acceleration signal of the horse during movement. The primary and secondary peaks of the acceleration magnitude are extracted based on a down-adjusted acceleration threshold. Each primary and secondary peak is identified as a single hoof-drop signal. Whenever a primary or secondary peak is extracted, signal fluctuations below the current primary or secondary peak are filtered out within a step counting window. When the triaxial acceleration signal is lost due to saturation, the horse's rotational envelope is calculated based on the geomagnetic vector, and the rotational envelope is time-aligned with the acceleration magnitude. Step compensation is performed based on the positive peak of the pitch angle within the rotational envelope. The final expression for the real-time gait frequency is: ,in For real-time cadence, The number of times the hoof lands corresponding to the main peak. This represents the number of times the secondary peak is hit. The number of hoof landings for step compensation; This is the step counting period;

[0030] The first and second step counting modules are used to adjust the step counting time window based on the movement speed according to the following formula: ,in Indicates the step counting time window. Indicates average stride length. This indicates the speed at which the horse moves.

[0031] The second step counting module is used to determine a threshold by linearly lowering the acceleration based on the velocity gradient according to the following formula: ,in The threshold for determining the reduced acceleration. The acceleration threshold corresponding to low-speed steady state. This is the horse gait attenuation coefficient. This represents the velocity gradient.

[0032] The gait re-admission condition refers to the velocity gradient exceeding a preset velocity gradient threshold.

[0033] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0034] 1. The method described in this invention establishes gait recombination admission conditions based on velocity gradients to distinguish between linear symmetrical gaits such as slow and fast steps and asymmetrical high-frequency gaits such as slow jumps and scrambles. For asymmetrical high-frequency gaits such as slow jumps and scrambles, the acceleration judgment threshold is dynamically lowered and the step counting time window is compressed based on the velocity gradient, thereby deconstructing the asymmetrical hoof landing signals that overlap in the time domain during high-speed gait recombination. This adapts to the peak characteristics formed by the dispersion of hoof impact energy under asymmetrical gait and identifies secondary feature points, thus effectively solving the problem of missed or miscounted gaits such as slow jumps and scrambles in traditional fixed threshold algorithms.

[0035] 2. The method described in this invention provides heartbeat support for step counting by using geomagnetic phase. When the accelerometer enters the signal saturation blind zone due to the huge instantaneous impact generated by the hoof landing, the geomagnetic phase is used to force logical alignment of the acceleration wave peak to ensure that the step counting pulse is not interrupted under severe impact.

[0036] 3. The method described in this invention utilizes the physical characteristics of horses' low frequency of heading changes and strong long-period regularity during track movement. It transforms geomagnetism from a general spatial navigation tool into a dedicated gait phase anchor point. When the gyroscope experiences pulse-like offset due to severe impact, a dedicated spatial coordinate system for gait is established through geomagnetic correction to compensate for the integral drift caused by severe impact. This ensures that step counting feature extraction is performed in a unified spatial dimension, greatly improving the sensing accuracy in non-stationary motion environments.

[0037] 4. The method described in this invention filters out invalid noise caused by high-frequency vibration of the sensor by setting a refractory period protection window for step counting through a step counting time window, and accurately retains step counting feature points that conform to the movement law of the horse's trunk, thereby improving the accuracy of hoof landing action recognition. Attached Figure Description

[0038] Figure 1 This is a flowchart of the method described in this invention.

[0039] Figure 2This diagram illustrates a comparison of the miscount rates of the method described in this invention and the traditional fixed threshold step counting method under different gait states.

[0040] Figure 3 This is a structural block diagram of the system described in this invention. Detailed Implementation

[0041] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0042] Example 1:

[0043] See Figure 1 An adaptive correction method for horse gait cycles is performed in the following steps:

[0044] S1. The sensor collects the three-axis acceleration signal, angular velocity signal and magnetic field vector of the horse in real time, and uses the magnetic field vector to correct the spatial coordinate system of the sensor.

[0045] Specifically, an integrated sensor system is installed in the horse's girth. This system includes a nine-axis motion sensor and a three-axis magnetometer. The nine-axis motion sensor comprises a three-axis accelerometer and a three-axis gyroscope. The three-axis accelerometer is used to collect the three-axis acceleration signals of the horse during movement in real time. A three-axis gyroscope is used to acquire the angular velocity signal of a horse in real time. A triaxial magnetometer is used to collect the triaxial magnetic field vector of the environment in which the horse is located. .

[0046] During intense horse racing training, the instantaneous impact of a horse's hoof striking the ground can easily cause pulse-like shifts in the angular velocity signals collected by the three-axis gyroscope, resulting in angular velocity drift. This invention utilizes a magnetic field vector to compensate for the integral drift of the angular velocity caused by the severe impact, obtaining a spatial coordinate system corrected by the geomagnetic vector. Using this spatial coordinate system as a reference, the spatial orientation of the three-axis acceleration and angular velocity signals output by the integrated sensor is corrected in real time, offsetting the integral drift error caused by the impact and ensuring the consistency of the spatial dimension for step counting feature extraction.

[0047] The process of establishing the spatial coordinate system is as follows: Gravity vector calibration (perpendicular to the Z-axis): During the horse's uniform linear motion, the gravity vector is extracted by low-pass filtering the three-axis acceleration signals. Using formulas Determine the vertical Z-axis of the coordinate system. Forward direction calibration (forward X-axis): Determine the horse's forward direction X-axis by integrating acceleration along the dominant motion frequency direction. Geomagnetic fusion correction: This is done using the magnetic field vector... With gravity vector The cross product operation determines the eastward vector in the horizontal plane. Using rotation matrices The integrated sensor coordinate system is transformed into a spatial coordinate system based on the gravity vector and corrected by the geomagnetic field.

[0048] S2. The horse's movement speed is collected in real time by the satellite positioning module integrated in the sensor, and the speed gradient is calculated based on the movement speed. The speed gradient is checked to see if it triggers the gait reorganization admission condition. If it is not triggered, proceed to S3. If it is triggered, proceed to S4.

[0049] Specifically, the movement speed is directly output by the satellite positioning module, and the speed gradient is the first-order difference between the movement speeds at adjacent sampling times, used to characterize the rate of change of the horse's speed over time. The gait reorganization admission condition refers to the speed gradient exceeding a preset speed gradient threshold. The speed gradient reflects the rate of change of the horse's speed over time. When it exceeds the preset speed gradient threshold, it indicates that the horse has entered an acceleration process and its gait is about to or is undergoing reorganization, transitioning from a linear symmetrical gait of walk and trot to an asymmetrical high-frequency gait such as a slow hop or scramble. At this time, the acceleration signal will exhibit distortion characteristics with overlapping peaks. The speed gradient threshold is obtained through statistical analysis of a large number of racehorse motion samples (critical speed gradients) transitioning from walk and trot to slow hop and scramble.

[0050] S3. Calculate the step counting time window based on the horse's movement speed, and count steps according to the acceleration threshold and the step counting time window.

[0051] Specifically, the step counting in S3 includes:

[0052] The acceleration magnitude is calculated based on the triaxial acceleration signal during horse movement. This acceleration magnitude is compared with an acceleration threshold, which corresponds to the acceleration threshold in low-speed steady state. If the acceleration magnitude exceeds the acceleration threshold and the time interval from the previous step pulse meets the step counting time window, it is determined to be a single hoof landing signal. The counter accumulates the hoof landing signals until the step counting cycle is reached. The final expression for the real-time gait frequency is: ,in For real-time cadence, To accumulate the number of hoof landings, The step counting cycle is preferably 3-5 seconds.

[0053] S4. Based on the velocity gradient, linearly lower the acceleration judgment threshold and calculate the step counting time window based on the movement speed. Then, count steps according to the lowered acceleration judgment threshold and the step counting time window.

[0054] Specifically, the step counting in S4 includes:

[0055] Calculation of acceleration magnitude based on triaxial acceleration signals during horse movement The acceleration magnitude is compared with the reduced acceleration threshold. When the acceleration magnitude exceeds the reduced acceleration threshold, the peak is determined to be a valid main peak and is used as a single hoof landing signal. Since the impact energy of the hoof landing is dispersed in the time domain after the horse enters a slow hop or a gait, the peak amplitude is significantly reduced compared to the low-speed steady state. If the acceleration threshold corresponding to the low-speed steady state is still used, a large number of real hoof landing main peaks will be missed due to insufficient amplitude. Therefore, it is necessary to linearly reduce the acceleration threshold based on the speed gradient to match the peak characteristics under the reconstructed gait.

[0056] The second derivative of the acceleration magnitude is obtained by taking the second derivative of the acceleration magnitude. When the second derivative signal has a local maximum and the maximum point is located in the downward interval of the main peak, it is further determined whether the acceleration magnitude at the corresponding moment of the maximum point exceeds the down-adjusted acceleration judgment threshold. If it exceeds the threshold, the maximum point is confirmed as a valid secondary peak and treated as a single gait signal. If it does not exceed the down-adjusted acceleration judgment threshold, the maximum point is determined to be an invalid fluctuation caused by sensor aftershocks or torso elastic vibration and is filtered out. The down-adjusted acceleration judgment threshold here serves to filter the amplitude of candidate secondary peaks identified by the second derivative method, avoiding misjudging weak vibrations not caused by gait as valid gait signals.

[0057] After each primary or secondary peak is extracted, signal fluctuations below 80% of the acceleration modulus corresponding to the current primary or secondary peak are filtered within the step counting time window to eliminate residual noise from the sensor caused by high-frequency vibrations, ensuring that each hoof stomp corresponds to only a single step pulse. When the triaxial acceleration signal is lost due to saturation, the horse's rotational envelope is calculated based on the geomagnetic vector, and the rotational envelope is time-aligned with the acceleration modulus. Step compensation is performed based on the positive peak of the pitch angle within the rotational envelope. The final expression for the real-time gait frequency is: ,in For real-time cadence, The number of times the hoof lands corresponding to the main peak. This represents the number of times the secondary peak is hit. The number of hoof landings for step compensation; This is the step counting cycle.

[0058] When a horse is in a trot or walk, its limb movements exhibit a clear linear rhythm, with independent and distinct acceleration signal peaks. However, when the horse enters a hop or trot state, its gait characteristics undergo a fundamental change. At this point, the horse's landing pattern changes from diagonally symmetrical to asymmetrical, high-frequency landing with three or four beats. In the acceleration amplitude signal from the sensor, this dense landing action manifests as multiple peaks highly overlapping and distorted in the time domain, indicating gait reorganization. This invention, during step counting, uses a down-adjusted acceleration threshold as a basis to identify overlapping primary and secondary peaks. A refractory period protection window is then set through a step counting time window to shield invalid signal fluctuations within the window, determining only valid landing actions outside the window. Simultaneously, the second derivative method of the acceleration modulus is used to identify secondary peaks in the descending range of the primary peak, incorporating both primary and secondary peaks into the step counting statistics, thus completing accurate step counting during the gait reorganization stage.

[0059] This invention first calculates the pitch and yaw angles during horse movement using magnetic field vectors. The periodic fluctuation components of these angles are then extracted using high-pass filtering to construct a rotating envelope of the geomagnetic vector. Simultaneously, the phase characteristics of this rotating envelope are time-aligned with the acceleration magnitude. When a saturation loss of the triaxial acceleration signal due to a high impact is detected (e.g., exceeding the sensor's range ±16g and lasting longer than 10ms), the positive peak of the pitch angle is extracted from the rotating envelope, and its timestamp is determined. Using this timestamp as a reference, and combined with a hysteresis constant set based on the horse's biomechanical characteristics, the effective step counting pulse's time position is deduced using a compensation formula, generating a compensated step counting signal. This compensated step counting signal is incorporated into the stride frequency statistics, completing the step counting compensation during the acceleration signal saturation phase. This step counting compensation ensures that the step counting pulse is not interrupted under severe impact. The compensation formula is: ,in This indicates the timestamp corresponding to the compensated step counting signal. The timestamp of the positive peak of the pitch angle. The physical fall time from the highest point of a horse's flight to the point where its hooves touch the ground.

[0060] Specifically, the step counting time window is adjusted based on the movement speed according to the following formula: ,in Indicates the step counting time window. Indicates average stride length. This indicates the speed at which the horse moves.

[0061] This invention calculates the stride counting time window based on the ratio of the horse's movement speed to its average stride length in slow hop and trotting gaits. This time window is synchronously compressed as the movement speed increases to match the time interval characteristics of high-frequency hoof landings, preventing two closely spaced hoof landings from being misjudged as a single hoof landing or a single noise point. The average stride length characteristic value can be obtained through statistical analysis of different racehorses during daily training in slow hop and trotting gaits.

[0062] Specifically, the threshold is determined by linearly reducing the acceleration based on the velocity gradient according to the following formula: ,in The threshold for determining the reduced acceleration. The acceleration threshold corresponding to low-speed steady state. This is the horse gait attenuation coefficient. This represents the velocity gradient.

[0063] This invention addresses the gait reorganization phenomenon in horses' slow hop and gait stride. Using the acceleration threshold under low-speed steady-state conditions as a benchmark, the threshold is linearly reduced based on the velocity gradient to adapt to the peak characteristics formed by the dispersion of impact energy during gait reorganization. The acceleration threshold under low-speed steady-state conditions can be determined by performing normality analysis on a large amount of acceleration peak distribution data from horses' slow and fast trots, taking the 70%-80% quantile of the acceleration amplitude as the acceleration threshold corresponding to the low-speed steady-state. The gait attenuation coefficient characterizes the degree of dispersion of the horse's landing impact force in the time domain; this coefficient typically ranges from 0.2 to 0.5.

[0064] Performance verification:

[0065] To verify the effectiveness of the method described in this invention, a simulation scene model containing a mixture of slow walking, fast walking, and rapid walking was constructed, and the step counting was performed using the method described in this invention. The results are as follows: Figure 2 As shown. Figure 2 This diagram illustrates the comparison of the miscount rates of the method described in this invention and the traditional fixed threshold step counting method under different gait conditions. As can be seen, in the low-speed range (0-15km / h), the accuracy of both methods is comparable. However, in the high-speed range (>30km / h) and at gait switching points, the miscount rate of the traditional fixed threshold step counting method increases exponentially because it cannot handle signal overlap. In contrast, the method described in this invention keeps the miscount rate below 5% across the entire speed range, which meets the accuracy requirements for step counting in professional horse racing training.

[0066] Example 2:

[0067] See Figure 3 An adaptive correction system for horse gait cycles includes a signal acquisition module, a gait judgment module, a first step counting module, and a second step counting module;

[0068] The signal acquisition module is used to acquire the three-axis acceleration signal, angular velocity signal and magnetic field vector of the horse in real time through the sensor, and to use the magnetic field vector to correct the spatial coordinate system of the sensor; specifically, the signal acquisition module is used to execute S1 in embodiment 1;

[0069] The gait judgment module is used to collect the horse's movement speed and speed gradient in real time, and detect whether the speed gradient triggers the gait reorganization admission condition. If it is not triggered, the first step counting module is used to count steps; if it is triggered, the second step counting module is used to count steps. The gait reorganization admission condition refers to the speed gradient exceeding a preset speed gradient threshold. The gait judgment module is used to execute S2 in embodiment 1.

[0070] The first step counting module is used to calculate the step counting time window based on the horse's movement speed, and to count steps according to the acceleration judgment threshold and the step counting time window. Specifically, the first step counting module counts steps according to the following steps: calculating the acceleration magnitude based on the horse's triaxial acceleration signal, determining whether the acceleration magnitude exceeds the acceleration judgment threshold, wherein the acceleration judgment threshold is the acceleration judgment threshold corresponding to low-speed steady state; if the acceleration magnitude exceeds the acceleration judgment threshold and the time interval from the previous step counting pulse meets the step counting time window, it is determined as a single hoof landing signal, and the counter accumulates the hoof landing signal until the step counting cycle is reached; the final expression for the real-time gait frequency is: ,in For real-time cadence, To accumulate the number of hoof landings, The step counting cycle is defined; the first step counting module is used to execute S3 in Example 1.

[0071] The second step counting module is used to linearly lower the acceleration judgment threshold based on the velocity gradient and calculate the step counting time window based on the horse's movement speed. Steps are counted according to the lowered acceleration judgment threshold and the step counting time window. Specifically, the second step counting module counts steps according to the following steps: calculating the acceleration magnitude based on the horse's triaxial acceleration signal; extracting the main peak and secondary peaks from the acceleration magnitude based on the lowered acceleration judgment threshold; classifying each main peak and secondary peak as a single hoof landing signal; filtering signal fluctuations below the current main peak or secondary peak within the step counting time window after each main peak or secondary peak is extracted; when the triaxial acceleration signal is saturated and lost, calculating the horse's rotation envelope based on the geomagnetic vector and aligning the rotation envelope with the acceleration magnitude in time; performing step counting compensation based on the positive peak of the pitch angle in the rotation envelope; finally, the expression for the real-time gait frequency is obtained as follows: ,in For real-time cadence, The number of times the hoof lands corresponding to the main peak. This represents the number of times the secondary peak is hit. The number of hoof landings for step compensation; This is the step counting period; specifically, the second step counting module is used to determine the threshold by linearly lowering the acceleration based on the velocity gradient according to the following formula: ,in The threshold for determining the reduced acceleration. The acceleration threshold corresponding to low-speed steady state. This is the horse gait attenuation coefficient. The velocity gradient is used; the second step counting module is used to execute S4 in Example 1.

[0072] The first and second step counting modules are used to adjust the step counting time window based on the movement speed according to the following formula: ,in Indicates the step counting time window. Indicates average stride length. This indicates the speed at which the horse moves.

[0073] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program goods. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0074] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0075] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0076] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An adaptive correction method for horse gait cycles, characterized in that: The adaptive correction method includes: S1. The sensor collects the three-axis acceleration signal, angular velocity signal and magnetic field vector of the horse in real time, and uses the magnetic field vector to correct the spatial coordinate system of the sensor. S2. Real-time acquisition of the horse's movement speed and speed gradient, detection of whether the speed gradient triggers the gait reorganization admission condition. If not triggered, proceed to S3; if triggered, proceed to S4. S3. Calculate the step counting time window based on the horse's movement speed, and count steps according to the acceleration threshold and the step counting time window. S4. Based on the velocity gradient, linearly lower the acceleration judgment threshold and calculate the step counting time window based on the horse's movement speed. Count steps according to the lowered acceleration judgment threshold and the step counting time window.

2. The adaptive correction method for horse gait cycle according to claim 1, characterized in that: The step counting in S3 includes: The acceleration magnitude is calculated based on the triaxial acceleration signal during horse movement. It is then determined whether the acceleration magnitude exceeds an acceleration threshold, which is the acceleration threshold corresponding to low-speed steady state. If the acceleration magnitude exceeds the acceleration threshold and the time interval from the previous step pulse meets the step counting time window, it is determined to be a single hoof landing signal. The counter accumulates the hoof landing signal until the step counting cycle is reached. The final expression for the real-time gait frequency is: ,in For real-time cadence, To accumulate the number of hoof landings, This is the step counting period; The step counting in S4 includes: The acceleration magnitude is calculated based on the triaxial acceleration signal of the horse during movement. The primary and secondary peaks of the acceleration magnitude are extracted based on a down-adjusted acceleration threshold. Each primary and secondary peak is identified as a single hoof-drop signal. Whenever a primary or secondary peak is extracted, signal fluctuations below the current primary or secondary peak are filtered out within a step counting window. When the triaxial acceleration signal is lost due to saturation, the horse's rotational envelope is calculated based on the geomagnetic vector, and the rotational envelope is time-aligned with the acceleration magnitude. Step compensation is performed based on the positive peak of the pitch angle within the rotational envelope. The final expression for the real-time gait frequency is: ,in For real-time cadence, The number of times the hoof lands corresponding to the main peak. This represents the number of times the secondary peak is hit. The number of hoof landings for step compensation; This is the step counting cycle.

3. The adaptive correction method for horse gait cycle according to claim 1 or 2, characterized in that: The step counting time window is adjusted based on the following formula, according to the movement speed: ,in Indicates the step counting time window. Indicates average stride length. This indicates the speed at which the horse moves.

4. The adaptive correction method for horse gait cycle according to claim 1 or 2, characterized in that: The threshold is determined by linearly reducing the acceleration based on the velocity gradient using the following formula: ,in The threshold for determining the reduced acceleration. The acceleration threshold corresponding to low-speed steady state. This is the horse gait attenuation coefficient. This represents the velocity gradient.

5. An adaptive correction method for horse gait cycle according to claim 1 or 2, characterized in that: The gait re-admission condition refers to the velocity gradient exceeding a preset velocity gradient threshold.

6. An adaptive correction system for horse gait cycles, characterized in that: The adaptive correction system includes a signal acquisition module, a gait judgment module, a first step counting module, and a second step counting module; The signal acquisition module is used to acquire the three-axis acceleration signal, angular velocity signal and magnetic field vector of the horse in real time through the sensor, and to use the magnetic field vector to correct the spatial coordinate system of the sensor. The gait judgment module is used to collect the horse's movement speed and speed gradient in real time, and detect whether the speed gradient triggers the gait reorganization admission condition. If it is not triggered, the first step counting module is used to count the steps; if it is triggered, the second step counting module is used to count the steps. The first step counting module is used to calculate the step counting time window based on the horse's movement speed, and to count steps according to the acceleration threshold and the step counting time window; The second step counting module is used to linearly lower the acceleration judgment threshold based on the speed gradient and calculate the step counting time window based on the horse's movement speed, and count steps according to the lowered acceleration judgment threshold and the step counting time window.

7. The adaptive correction system for horse gait cycle according to claim 6, characterized in that: The first step counting module counts steps according to the following steps: The acceleration magnitude is calculated based on the triaxial acceleration signal of the horse during movement, and it is determined whether the acceleration magnitude exceeds the acceleration determination threshold. The acceleration determination threshold is the acceleration determination threshold corresponding to the low-speed steady state. If the acceleration magnitude exceeds the acceleration judgment threshold and the time interval from the previous step counting pulse meets the step counting time window, it is judged as a single footfall signal. The counter accumulates the footfall signal until the step counting cycle is reached; the final expression for the real-time step frequency is: ,in For real-time cadence, To accumulate the number of hoof landings, This is the step counting period; The second step counting module counts steps according to the following steps: The acceleration magnitude is calculated based on the triaxial acceleration signal of the horse during movement. The primary and secondary peaks of the acceleration magnitude are extracted based on a down-adjusted acceleration threshold. Each primary and secondary peak is identified as a single hoof-drop signal. Whenever a primary or secondary peak is extracted, signal fluctuations below the current primary or secondary peak are filtered out within a step counting window. When the triaxial acceleration signal is lost due to saturation, the horse's rotational envelope is calculated based on the geomagnetic vector, and the rotational envelope is time-aligned with the acceleration magnitude. Step compensation is performed based on the positive peak of the pitch angle within the rotational envelope. The final expression for the real-time gait frequency is: ,in For real-time cadence, The number of times the hoof lands corresponding to the main peak. This represents the number of times the secondary peak is hit. The number of hoof landings for step compensation; This is the step counting cycle.

8. An adaptive correction system for horse gait cycles according to claim 6 or 7, characterized in that: The first and second step counting modules are used to adjust the step counting time window based on the movement speed according to the following formula: ,in Indicates the step counting time window. Indicates average stride length. This indicates the speed at which the horse moves.

9. An adaptive correction system for horse gait cycles according to claim 6 or 7, characterized in that: The second step counting module is used to determine a threshold by linearly lowering the acceleration based on the velocity gradient according to the following formula: ,in The threshold for determining the reduced acceleration. The acceleration threshold corresponding to low-speed steady state. This is the horse gait attenuation coefficient. This represents the velocity gradient.

10. An adaptive correction system for horse gait cycles according to claim 6 or 7, characterized in that: The gait re-admission condition refers to the velocity gradient exceeding a preset velocity gradient threshold.

Images