CN-122004865-A - Agricultural machinery driving fatigue recognition method based on palm rate change of machinery

CN122004865ACN 122004865 ACN122004865 ACN 122004865ACN-122004865-A

Abstract

An agricultural machinery driving fatigue recognition method based on the change of the palm rate of an agricultural machinery belongs to the technical field of intelligent monitoring and man-machine engineering intersection of agricultural machinery. The invention aims at the problems that the existing physiological signal detection method of driving fatigue is difficult in feature extraction and has poor universality because of individual difference. The method comprises the steps of extracting features based on heart rate sequences to obtain time domain heart rate features and nonlinear heart rate features, simultaneously calculating based on power spectrum density to obtain frequency domain heart rate features, processing to obtain corrected time domain fatigue features, corrected frequency domain fatigue features, corrected nonlinear fatigue complexity features and corrected multi-time-scale driving fatigue cooperative features, calculating multi-feature coupling fatigue energy, calculating fatigue accumulation response and fatigue discrimination probability values, calculating fatigue strength indexes and self-adaptive fatigue classification thresholds, and finally obtaining fatigue grade discrimination results. The invention is used for recognizing the fatigue state of the agricultural machinery hand.

Inventors

WANG YIJIA
WANG FEIFAN
WEN NUAN
AN JIAN
WANG NAIHUI
XIE ZHENZHEN

Assignees

东北农业大学

Dates

Publication Date: 20260512
Application Date: 20260127

Claims (10)

1. An agricultural machinery driving fatigue recognition method based on the change of the palm rate of an engine is characterized by comprising the following steps, Acquiring a discrete electrocardiosignal sequence of a manipulator, and preprocessing to obtain a heart rate sequence; Performing feature extraction based on the heart rate sequence to obtain time domain heart rate features and nonlinear heart rate features; Calculating to obtain a time domain fatigue sensitive heart rate characteristic, a frequency domain fatigue unbalance characteristic and a nonlinear fatigue complexity characteristic based on the time domain heart rate characteristic, the frequency domain heart rate characteristic and the nonlinear heart rate characteristic; calculating to obtain a multi-time-scale driving fatigue cooperative characteristic based on the time-domain fatigue sensitive heart rate characteristic and the frequency-domain fatigue unbalance characteristic; correcting the time domain fatigue sensitive heart rate characteristic, the frequency domain fatigue unbalance characteristic and the nonlinear fatigue complexity characteristic to obtain corrected time domain fatigue characteristic, corrected frequency domain fatigue characteristic and corrected nonlinear fatigue complexity characteristic; Calculating multi-feature coupling fatigue energy based on the corrected time domain fatigue feature, the corrected frequency domain fatigue feature and the corrected nonlinear fatigue complexity feature, calculating fatigue accumulation response, and calculating a fatigue discrimination probability value by combining the corrected multi-time-scale driving fatigue cooperative feature; Calculating a fatigue strength index according to the fatigue discrimination probability value and the corrected multi-time-scale driving fatigue cooperative characteristic, and calculating a self-adaptive fatigue grading threshold value based on the corrected multi-time-scale driving fatigue cooperative characteristic; And combining the fatigue strength index and the adaptive fatigue grading threshold to obtain a fatigue grade judging result.
2. The method for recognizing the fatigue of the driving of the agricultural machinery based on the change of the palm rate of the machinery according to claim 1, wherein, Representing the discrete electrocardiosignal sequence of the manipulator The pretreatment process comprises the following steps: Discrete electrocardiosignal sequence for manipulator according to selected frequency band Frequency band constraint is carried out: , In the middle of For the electrocardiosignals after the frequency band is constrained, For the number of the electrocardiosignals after the frequency band is constrained, The number of the electrocardiosignals is discrete by the manipulator, Is a band-pass filter operator; Electrocardiosignals after frequency band constraint Performing adaptive smoothing and differential operation to obtain enhanced electrocardiosignal : , In the middle of To take the following measures Is the locally smoothed window length of the center, In order to calculate the sum of the variables, ; A wavelet denoising operator; , In the middle of For a base smoothing of the window length, To enhance the tuning parameters; Based on enhanced post-electrocardiosignal Construction of cardiac significance sequences : , Sequence of significance to heart beat Detecting peak value under amplitude threshold constraint and physiological normal RR interval range constraint to obtain heart beat position sequence : , In the middle of For the index of the kth heart beat position, For the peak detection operator, In order to be a significance amplitude threshold value, Is the minimum RR interval; calculating corrected RR intervals based on adjacent heart beat positions : , In the middle of Is the local RR interval median value; thereby calculating and obtaining heart rate sequence : 。
3. The method for recognizing the fatigue of the driving of the agricultural machinery based on the change of the palm rate of the machinery according to claim 2, wherein, The calculation method of the time domain heart rate characteristic, the frequency domain heart rate characteristic and the nonlinear heart rate characteristic comprises the following steps: Setting a time window length for heart rate analysis Then at the time window length Heart rate subsequence within The method comprises the following steps: , ; time domain heart rate characterization The method comprises the following steps: , In the middle of For the length of the time window The number of internal heartbeats, For the length of the time window An internal average heart rate; Frequency domain heart rate characterization The method comprises the following steps: , In the middle of In the range of the low-frequency band, Is in the full frequency range, As a function of the frequency variation, For the length of the time window Power spectral density of the internal heart rate sequence; Nonlinear heart rate characteristics The method comprises the following steps: 。
4. The method for recognizing the fatigue of the driving of the agricultural machinery based on the change of the palm rate of the machinery according to claim 3, wherein, The calculation method of the time domain fatigue sensitive heart rate characteristic, the frequency domain fatigue unbalance characteristic and the nonlinear fatigue complexity characteristic comprises the following steps: time domain fatigue sensitive heart rate characteristics The method comprises the following steps: , In the middle of Is an integral variable; frequency domain fatigue imbalance characterization The method comprises the following steps: ; Nonlinear fatigue complexity characterization The method comprises the following steps: , , In the middle of Is the difference between adjacent heart beat and heart rate.
5. The method for recognizing fatigue of driving of an agricultural machine based on the change of the palm rate of the machine according to claim 4, wherein, Multi-time-scale driving fatigue cooperative characteristic is expressed as : , In the middle of Time domain fatigue sensitive heart rate characterization for various time scales Is used for the average value of (a), Frequency domain fatigue imbalance characterization for each time scale Average value of (2); 。
6. the method for recognizing fatigue of driving of an agricultural machine based on the change of the palm rate of the machine according to claim 5, wherein, The calculation method of the corrected time domain fatigue characteristics, the corrected frequency domain fatigue characteristics and the corrected nonlinear fatigue complexity characteristics comprises the following steps: corrected time domain fatigue characteristics The method comprises the following steps: , In the middle of For the individual heart rate baseline mapping factor, A nonlinear correction factor for the environmental load; , , In the middle of For the average heart rate of the robot in a long-term working state, For the reference heart rate value, Is an environmental load intensity index; Post-correction frequency domain fatigue characteristics The method comprises the following steps: ; corrected nonlinear fatigue complexity characteristics The method comprises the following steps: 。
7. the method for recognizing fatigue of driving of an agricultural machine based on the change of the palm rate of the machine according to claim 6, wherein, Multi-time-stamp driving fatigue cooperative characteristic after correction The method comprises the following steps: 。
8. The method for recognizing fatigue of driving of an agricultural machine based on the change of the palm rate of the machine according to claim 7, wherein, The multi-feature coupling fatigue energy calculating method comprises the following steps: , In the middle of For the coupling of fatigue energy for multiple features, For the time-domain fatigue feature map quantity, For the frequency domain fatigue characteristic mapping quantity, Mapping the nonlinear fatigue characteristics; ; cumulative response to fatigue The method comprises the following steps: , In the middle of Is an integral variable; Fatigue discrimination probability value The method comprises the following steps: 。
9. the method for recognizing fatigue of driving of an agricultural machine based on the change of the palm rate of the machine according to claim 8, wherein, Fatigue strength index The method comprises the following steps: , In the middle of An extremely small positive number for preventing the denominator from being zero; adaptive fatigue classification threshold The method comprises the following steps: , In the middle of And judging the threshold value based on the fatigue.
10. The method for recognizing the fatigue of the driving of the agricultural machinery based on the change of the palm rate of the machinery according to claim 9, wherein, Fatigue grade determination result The method comprises the following steps: 。

Description

Agricultural machinery driving fatigue recognition method based on palm rate change of machinery Technical Field The invention relates to an agricultural machinery driving fatigue recognition method based on the change of the palm rate of an engine, belonging to the technical field of intelligent monitoring and man-machine engineering intersection of agricultural machinery. Background With the continuous improvement of agricultural mechanization and intelligence, agricultural machinery has become the core equipment of modern agricultural production. When agricultural machinery such as a tractor and a combine harvester is operated in the field, it is generally necessary to perform a long-time, high-load and continuous operation task. In the process, the manipulator not only bears high-strength mechanical vibration, noise and dust invasion, but also faces multiple adverse factors such as monotonous and repeated operation, severe changes of day and night illumination, operation psychological pressure and the like, and the rapid accumulation of rational and psychological fatigue is extremely easy to cause. Under the fatigue state, the response of the machine hand is slow, the misoperation is increased, and the safety accidents such as mechanical overturning, collision and the like can be possibly caused when the machine is serious, so that the personal safety is threatened, and the operation efficiency and the agricultural production benefit are directly influenced. At present, the driving fatigue detection technology mainly comprises three types, namely a method based on vehicle operation parameters, a method based on facial visual characteristics of a driver and a method based on physiological signals. The vehicle state method recognizes fatigue by monitoring steering wheel rotation angle, lane departure, vehicle speed fluctuation and the like, but under the conditions of low operation speed, irregular running track and large topography fluctuation of the agricultural machinery, the vehicle parameters are obviously interfered by operation content and ground surface conditions, and the false detection rate is high. The facial image method relies on a camera to capture behavior characteristics such as blink frequency, pupil state, yawning and the like, but the phenomena of uneven illumination, dust adhesion, shielding objects such as caps and masks worn by drivers and the like in farmland environments are ubiquitous, and the image quality and the recognition stability are seriously affected. Physiological signal method (such as brain electricity, electrocardio and skin electricity) directly reflects the state of autonomous nervous system of human body, has strong objectivity, but under the severe working condition of strong vibration and high noise of farm machinery, physiological signals are easy to suffer from motion artifact and electromagnetic interference, signal quality is reduced, and feature extraction is difficult. The agricultural machinery operation has remarkable scene specificity, namely, the vibration intensity is high, the frequency spectrum is complex, and the electrocardiosignals and the vibration noise are difficult to effectively separate by the traditional filtering method; the fatigue model with a single time scale is difficult to accurately describe, and the fatigue model has obvious influence on individual differences and environmental loads, different heart rate baselines and physical conditions of different manipulators, different physiological loads caused by different fields and different operation links and poor applicability of a general fatigue threshold. Therefore, the existing fatigue detection method faces the technical bottlenecks of insufficient adaptability, poor stability and large interference between individuals and the environment in the agricultural machinery operation scene. Disclosure of Invention Aiming at the problems that the existing physiological signal detection method of driving fatigue is difficult in feature extraction and the universality is poor due to individual differences, the invention provides an agricultural machinery driving fatigue identification method based on the change of the palm rate of an engine. The invention relates to an agricultural machinery driving fatigue recognition method based on the change of the palm rate of an engine, which comprises the following steps, Acquiring a discrete electrocardiosignal sequence of a manipulator, and preprocessing to obtain a heart rate sequence; Performing feature extraction based on the heart rate sequence to obtain time domain heart rate features and nonlinear heart rate features; Calculating to obtain a time domain fatigue sensitive heart rate characteristic, a frequency domain fatigue unbalance characteristic and a nonlinear fatigue complexity characteristic based on the time domain heart rate characteristic, the frequency domain heart rate characteristic and the nonlinear heart rate characteristic; c