CN108992072B - A gait event-based human lower limb movement intention recognition method - Google Patents

Abstract

The invention discloses a human body intention identification method based on gait event detection. The method comprises the steps of firstly defining the gait event definition of the lower limbs of the human body, then detecting the gait event in real time, then obtaining the real-time step frequency and the stride based on the gait event, finally adjusting the standard joint corner track according to the real-time step frequency and the stride, and generating the reference joint corner track for synchronizing the step frequency and the stride. According to the invention, joint corner data can be acquired in real time through an inertia detection unit (IMU) installed on the lower limb, and a joint corner reference track for synchronizing the step frequency and the stride is obtained according to the acquired data and is used as a tracking track for motion control, so that the synchronous human gait of the lower limb exoskeleton is finally realized.

Description

A gait event-based human lower limb movement intention recognition method

technical field

The invention relates to a method for recognizing human lower limb movement intention based on gait events.

Background technique

With the continuous development of electromechanical technology, the research of robotic systems has become a hot topic at this stage. One of the wearable lower limb exoskeleton robots has made staged progress and has a wide range of applications in the military and medical fields. application.

Exoskeleton robot recognition of the motion intention of human lower limbs is one of the key technologies to realize the intelligentization of exoskeleton robots. At this stage, there are mainly two methods to realize lower limb motion intention recognition. One is to use human body electrical signals to identify human body motion intentions, such as electromyography (EMG) or brain waves (EOG); the other is to detect human-computer interaction force. to identify human motion intentions.

SUMMARY OF THE INVENTION

Aiming at the deficiencies of the prior art, the present invention provides a method for recognizing human lower limb movement intention based on gait events.

The invention firstly detects the real-time gait parameters (step frequency and stride length) of the lower limbs of the human body, then adjusts the standard rotation angle trajectory according to the real-time cadence frequency and stride length, and then obtains the joint rotation angle control target trajectory synchronizing the human body's cadence frequency and stride length, and finally The target trajectory is used as the tracking target trajectory of the control system for motion control, so that the lower limb exoskeleton can move synchronously with the human body's gait, that is, the recognition of the motion intention of the lower limbs of the human body is realized.

The specific content of the technical solution of the present invention is as follows:

(1) Definition of human lower limb gait events.

(2) Real-time detection algorithm of gait events.

(3) Real-time stride frequency and stride length are obtained based on gait events.

(4) Adjust the standard joint angle trajectory according to the real-time cadence and stride, and generate the reference joint angle trajectory of the synchronized cadence and stride.

The beneficial effects of the present invention are as follows: a brand-new lower limb motion intention recognition method is proposed, which can acquire joint rotation angle data in real time through an inertial detection unit (IMU) installed on the lower limb, and obtain synchronous cadence and cadence according to the measured data. The joint rotation angle reference trajectory of the width is used as the tracking trajectory of the motion control, and finally the human gait synchronization of the lower limb exoskeleton is realized.

Description of drawings

Figure 1 is a block diagram of the overall method;

Figure 2 is a gait event definition diagram.

Detailed ways

The overall scheme of the present invention is mainly divided into three steps, wherein the first step is real-time gait event detection, the second step is real-time gait frequency and stride detection, and the last step is to adjust the standard joint corner trajectory curve and generate a reference corner trajectory. Specifically as shown in Figure 1.

(1) Gait event detection

The present invention divides a complete gait cycle into four gait phases, including the initial stance (ISt), mid-stance (MSt), end-stance (TSt) and swing phase (Sw). The start of each gait phase is defined as a gait event, including the contact initiation event (IC), the thigh erection event (TE), the toe pre-off event (pTO) and the toe off the ground event (TO), as shown in Figure 2 shown.

According to the definition of gait events and the experimental results, it can be found that there is a certain mapping relationship between gait events and the rotation angle of the lower extremity hip joint. The specific rules are shown in Table 1.

Table 1 is the regularity of gait events and hip joint rotation

Gait events IC TE pTO TO Hip Rotation maximum value about zero minimum about zero

The data acquisition card collects the IMU real-time hip joint angle data of the lower limbs of the human body, and writes a program according to the above rules to detect each gait event in each gait cycle in real time.

(2) Step frequency and stride length algorithm

2.1 Step frequency algorithm

According to the result of gait event detection, the time interval of adjacent same gait events can be obtained, and then the calculation formula of gait cycle T _s can be derived as follows:

where t _n and t _n-1 represent the time points of the nth and n-1th gait events, respectively. In order to improve the accuracy of gait cycle calculation, the four gait events defined in the gait cycle are used to calculate the gait cycle and take the average value. The specific formula is as follows:

和

分别表示第n次和第n-1次第i个步态事件发生的时间点，ω_s表示步态频率，k等于4。得到了步态周期后，再定义步态周期的相位角为：in the formula

and

represent the time points of the nth and n-1th ith gait events, respectively, ωs represents the gait frequency, and _k is equal to 4. After obtaining the gait cycle, define the phase angle of the gait cycle as:

where φ _s represents the gait phase angle function, which ranges from 0 to 2π, and then defines the gait progress percentage s according to the phase angle as:

2.2 Stride algorithm

In this design, the stride is defined as the distance between the center of mass of the feet when the contact start event or the toe pre-ground event occurs. According to the above definition and the simplified model of lower limbs, the calculation formula of stride length can be obtained as follows:

L=l ₃ sinθ ₃ +l ₄ sinθ ₄ -(l ₁ sinθ ₁ +l ₂ sinθ ₂ )

在一个完整的步态周期中，存在着两个步态事件可以用来计算步幅，为了提高计算结果的精度，可将步幅计算公式表示如下：where θ ₁ and θ ₂ represent the knee joint rotation angle, θ ₃ and θ ₄ represent the hip joint rotation angle,

In a complete gait cycle, there are two gait events that can be used to calculate the stride length. In order to improve the accuracy of the calculation results, the stride length calculation formula can be expressed as follows:

和

表示当IC(i＝1)和pTO(i＝2)步态事件发生时髋关节转角值，

和

表示当IC(i＝1)和pTO(i＝2)步态事件发生时膝关节转角值。in the formula

and

represents the value of hip joint rotation when IC(i=1) and pTO(i=2) gait events occur,

and

Indicates the knee joint rotation value when IC (i=1) and pTO (i=2) gait events occur.

(3) Joint rotation angle trajectory adjustment algorithm

Exoskeleton reference joint tracking trajectories are generated based on standard joint trajectories, including hip and knee joints. The standard joint rotation trajectory is obtained by referring to the clinical gait analysis database (CGA) or experimental results. In order to realize the synchronized lower limb motion intention of the exoskeleton robot, the standard joint angle trajectory needs to adjust the period and amplitude according to the stride frequency and stride detected in real time.

3.1 Step frequency adjustment algorithm

Assuming that the standard hip and knee rotation curve functions are f _h (s) and f _k (s), respectively, the s variable represents the percentage of gait progress, and s is a time-varying parameter, which is specifically defined as follows:

where φ _s (t) represents the gait phase angle function, which can be obtained in the gait detection algorithm. Then, the joint angle function of the synchronous cadence can be derived as follows:

f _h (t)=f _h (s(t))

f _k (t)=f _k (s(t))

3.2 Stride adjustment algorithm

Assuming that the stride of the standard gait is L _s , and the result of stride detection is L _r , the stride adjustment parameter K can be defined as:

According to the detection result of gait event, the percentage value of gait process when TO event occurs can be obtained. The specific calculation formula is:

In the formula, t _TO represents the time point of the TO event, t _TE represents the time point of the TE event, and T _s represents the gait cycle. Then adjust the standard trajectory function as follows:

where f(s) represents the standard joint angle function, Θ ₁ is the stride stretch interval, Θ ₁ ∈(0,s _TO ), Θ ₂ is the stride contraction interval, Θ ₂ ∈(s _TO ,100%). Finally, when applied to the step frequency adjustment algorithm, the formula can be derived as:

^fn (t)= ^fn (s(t))

where f ⁿ (t) represents the reference corner trajectory function, that is, the corner trajectory curve that synchronizes the human cadence and stride length.

To sum up, the present invention designs a human intention recognition method based on gait event detection. This method uses four inertial detection units (IMUs) to detect the joint rotation angles of the lower limbs (hip and knee joints) in real time. In terms of the complexity of the algorithm, the algorithm designed by the present invention is also relatively simple, which reduces the amount of calculation.

Claims (3)

1. a human lower limb movement intention recognition method based on gait event, is characterized in that, comprises the following steps: (1) Define human lower limb gait events; The start of each gait phase is a gait event, including contact start event, thigh upright event, toe pre-lifting event, and toe-lifting event, these events have a corresponding mapping relationship with hip joint rotation; (2) Real-time detection of gait events; Collect real-time hip joint rotation angle data of human lower limbs, and detect each gait event in each gait cycle according to the above mapping relationship. Each gait cycle is divided into four gait phases, including initial support, middle support, end of support and swing; (3) Obtain real-time stride frequency and stride length based on gait events; The process of obtaining the real-time cadence is as follows: According to the result of gait event detection, the time interval between adjacent same gait events is obtained, and the gait cycle T _s is deduced:

where t _n and t _n-1 represent the time points of the nth and n-1th gait events, respectively, and _ωs represents the gait frequency; After obtaining the gait cycle, define the phase angle of the gait cycle as:

where φ _S is the gait phase angle function, which ranges from 0 to 2π, and then defines the gait process percentage s according to the phase angle as:

The process of getting the real-time stride is: The stride is defined as the distance between the centers of mass of the two feet when the contact start event or the toe pre-lifting event occurs, and the calculation formula of the stride is as follows: L=l ₃ sinθ ₃ +l ₄ sinθ ₄ -(l ₁ sinθ ₁ +l ₂ sinθ ₂ ) where θ ₁ and θ ₂ represent the knee joint rotation angle, θ ₃ and θ ₄ represent the hip joint rotation angle,

(4) Adjust the standard joint angle trajectory according to the real-time cadence and stride, and generate the reference joint angle trajectory of synchronous cadence and stride; Cadence adjustment: Assuming that the standard hip and knee joint rotation curve functions are f _h (s) and f _k (s), respectively, the joint rotation function of the synchronous cadence is as follows: f _h (t)=f _h (s(t)) f _k (t)=f _k (s(t)) Stride Adjustment: Assuming that the stride of the standard gait is L _s and the result of stride detection is L _r , the stride adjustment parameter K is defined as:

According to the gait event detection results, the gait process percentage value when the toe-off event occurs is obtained:

In the formula, t _TO represents the time point of the toe off the ground event, and t _TE represents the time point of the thigh upright event; then adjust the standard trajectory function as follows:

where f(s) represents the standard joint angle function, Θ ₁ is the stride stretch interval, Θ ₁ ∈(0, s _TO ), Θ ₂ is the stride contraction interval, Θ ₂ ∈(s _TO , 100%), Finally applied to the cadence adjustment, we get: ^fn (t)= ^fn (s(t)) where f ⁿ (t) represents the reference corner trajectory function, that is, the corner trajectory curve that synchronizes the human cadence and stride length. 2. a kind of human lower limb motion intention recognition method based on gait event according to claim 1 is characterized in that: in order to improve the accuracy of gait cycle calculation, the four gait events defined in the gait cycle are all Used to calculate the gait cycle and take the average, the specific formula is as follows:

in the formula

and

represent the time points of the nth and n-1th ith gait events, respectively. 3. a kind of human lower limb motion intention recognition method based on gait event according to claim 1 is characterized in that: in a complete gait cycle, there are two gait events that can be used to calculate stride length, In order to improve the accuracy of the calculation results, the stride calculation formula is expressed as follows:

in the formula

and

Represents the value of the hip joint rotation when the contact start event and the toe pre-ground event occur,

and

Indicates the knee joint rotation value when the contact start event and the toe pre-ground event occur.

Images