Continuous joint angle estimation based on a surface electromyography (sEMG) signal can be used to improve the man-machine coordination performance of the exoskeleton. In this study, we proposed a time-advanced feature and utilized long short-term memory (LSTM) with a root mean square (RMS) feature and its time-advanced feature (RMSTAF; collectively referred to as RRTAF) of sEMG to estimate the knee joint angle. To evaluate the effect of joint angle estimation, we used root mean square error (RMSE) and cross-correlation coefficient ρ between the estimated angle and actual angle. We also compared three methods (i.e., LSTM using RMS, BPNN (back propagation neural network) using RRTAF, and BPNN using RMS) with LSTM using RRTAF to highlight its good performance. Five healthy subjects participated in the experiment and their eight muscle (i.e., rectus femoris (RF), biceps femoris (BF), semitendinosus (ST), gracilis (GC), semimembranosus (SM), sartorius (SR), medial gastrocnemius (MG), and tibialis anterior (TA)) sEMG signals were taken as algorithm inputs. Moreover, the knee joint angles were used as target values. The experimental results showed that, compared with LSTM using RMS, BPNN using RRTAF, and BPNN using RMS, the average RMSE values of LSTM using RRTAF were respectively reduced by 8.57%, 46.62%, and 68.69%, whereas the average ρ values were respectively increased by 0.31%, 4.15%, and 18.35%. The results demonstrated that LSTM using RRTAF, which contained the time-advanced feature, had better performance for estimating the knee joint motion.

Hikers and soldiers usually walk up and down slopes with a load carriage, causing injuries of the musculoskeletal system, especially during a prolonged load journey. The slope walking has been reported to lead to higher leg extensor muscle activities and joint moments. However, most of the studies investigated muscle activities or joint moments during slope walking without load carriage or only investigated the joint moment changes and muscle activities with load carriages during level walking. Whether the muscle activation such as the signal amplitude is influenced by the mixed factor of loads and grades and whether the influence of the degrees of loads and grades on different muscles are equal have not yet been investigated. To explore the effects of backpack loads on leg muscle activation during slope walking, ten young male participants walked at 1.11 m/s on a treadmill with different backpack loads (load masses: 0, 10, 20, and 30 kg) during slope walking (grade: 0, 3, 5, and 10°). Leg muscles, including the gluteus maximus (GM), rectus femoris (RF), hamstrings (HA), anterior tibialis (AT), and medial gastrocnemius (GA), were recorded during walking. The hip, knee, and ankle extensor muscle activations increased during the slope walking, and the hip muscles increased most among hip, knee, and ankle muscles (GM and HA increased by 46% to 207% and 110% to 226%, respectively, during walking steeper than 10° across all load masses (GM: p = 1.32 × 10⁻⁸ and HA: p = 2.33 × 10⁻¹⁶)). Muscle activation increased pronouncedly with loads, and the knee extensor muscles increased greater than the hip and ankle muscles (RF increased by 104% to 172% with a load mass greater than 30 kg across all grades (RF: p = 8.86 × 10⁻⁷)). The results in our study imply that the hip and knee muscles play an important role during slope walking with loads. The hip and knee extension movements during slope walking should be considerably assisted to lower the muscle activations, which will be useful for designing assistant devices, such as exoskeleton robots, to enhance hikers’ and soldiers’ walking abilities.

Numerous robots have been widely used to deliver rehabilitative training for hemiplegic patients to improve their functional ability. Because of the complexity and diversity of upper limb motion, customization of training patterns is one key factor during upper limb rehabilitation training. Most of the current rehabilitation robots cannot intelligently provide adaptive training parameters, and they have not been widely used in clinical rehabilitation. This article proposes a new end-effector upper limb rehabilitation robot, which is a two-link robotic arm with two active degrees of freedom. This work investigated the kinematics and dynamics of the robot system, the control system, and the realization of different rehabilitation therapies. We also explored the influence of constraint in rehabilitation therapies on interaction force and muscle activation. The deviation of the trajectory of the end effector and the required trajectory was less than 1 mm during the tasks, which demonstrated the movement accuracy of the robot. Besides, results also demonstrated the constraint exerted by the robot provided benefits for hemiplegic patients by changing muscle activation in the way similar to the movement pattern of the healthy subjects, which indicated that the robot can improve the patient’s functional ability by training the normal movement pattern.

This paper proposes a hierarchical support vector machine recognition algorithm based on a finite state machine (FSM-HSVM) to accurately and reliably recognize the locomotion mode recognition of an exoskeleton robot. As input signals, this method utilizes the angle information of the hip joint and knee joint collected by inertial sensing units (IMUs) on the thighs and shanks of the exoskeleton and the plantar pressure information collected by force sensitive resistors (FSRs) are used as input signals. This method establishes a framework for mode transition by combining the finite state machine (FSM) with the common locomotion modes. The hierarchical support vector machine (HSVM) recognition model is then tightly integrated with the mode transition framework to recognize five typical locomotion modes and eight locomotion mode transitions in real-time. The algorithm not only reduces the abrupt change in the recognition of locomotion mode, but also significantly improves the recognition efficiency. To evaluate recognition performance, separate experiments are conducted on six subjects. According to the results, the average accuracy of all motion modes is 97.106% ± 0.955%, and the average recognition delay rate is only 25.017% ± 6.074%. This method has the benefits of a small calculation amount and high recognition efficiency, and it can be applied extensively in the field of robotics.

In the research and development of a passive exoskeleton, the body–exoskeleton coupling mode is a key point to reduce the interaction force and realize the efficient assistance of the exoskeleton. The purpose of this paper was to explore a cooperative movement mode between human and passive exoskeleton for reducing the body–exoskeleton interaction force. Firstly, through the research of the body–exoskeleton interactive mode, we analyzed the kinematic and dynamic constraint of the exoskeleton and established a dynamic model of the body–exoskeleton system. On this basis, the characteristic of the body–exoskeleton interaction force was analyzed; then, we put forward a mode that uses human gravity and load weight to maintain the stability of the exoskeleton’s movement to achieve the goal of reducing the interaction force. Based on the human–exoskeleton integrated mode, we constructed a mechanical model and simulated the change in interaction force in this mode; the simulation results showed that the interaction force at the lower leg was 98.5% less than that of the pure mechanical exoskeleton. Finally, we developed a prototype that was made of plastic parts and finished the experiment by walking with a load of 30 kg. The experimental results showed that this mode reduced the body–exoskeleton interaction force by 65.1%, which verified the effectiveness of the body–exoskeleton coupling mode preliminarily. The research results provided a new analytical approach for the design of a passive exoskeleton, and its improvement effect could be extended from the lower leg of the body–exoskeleton to the thigh or trunk, and guide the design of a passive exoskeleton.

Real-time continuous locomotion mode recognition and seamless timely transition detection is critical for the exoskeleton robot. This study aims to present a comprehensive and innovative framework for locomotion mode recognition and transition prediction, exclusively utilizing inertial measurement unit (IMU) signals from the exoskeleton. In this framework, a CNN-BiLSTM model was developed and trained to be the classifier and a novel majority filter was designed to reduce the transition misjudgment rate. Moreover, a comprehensive evaluation system encompassing eight dimensions for the classifier, incorporating evaluation metrics specifically for transition misjudgment, was proposed. We collected locomotion motion data from six subjects wearing a rigid exoskeleton robot using six IMU sensors on the exoskeleton. The proposed method achieves a high level of recognition accuracy, with an overall average of 99.58% for the five steady locomotion modes (level ground walking (LG), stair ascent/descent (SA/SD), and ramp ascent/descent (RA/RD)) across six subjects following the transition decision. All transitions are recognizable, and the majority can be predicted in advance, with an average prediction time of 353ms. Furthermore, the implementation of majority filter resulted in an average 87.04% reduction in the transition misjudgment rate among six subjects, thereby decreasing the average transition misjudgment rate to 4.82%. Finally, the model was tested on a Jetson Nano to verify its real-time performance. The results presented above were obtained under the condition where either leg could function as the first transition leg and revealed that the developed system was capable of achieving precise locomotion mode recognition and timely transition prediction, with high real-time performance.