Contact State Recognition Research Articles

Contact State Recognition for Dual Peg-in-Hole Assembly of Tightly Coupled Dual Manipulator

Contact state recognition is a critical technology for enhancing the robustness of robotic assembly tasks. There have been many studies on contact state recognition for single-manipulator, single peg-in-hole assembly tasks. However, as the number of pegs and holes increases, the contact state becomes significantly more complex. Additionally, when a tightly coupled multi-manipulator is required, the estimation errors in the contact forces between pegs and holes make contact state recognition challenging. The current state recognition methods have not been tested in such tasks. This paper tested Support Vector Machine (SVM) and several neural network models on these tasks and analyzed the recognition accuracy, precision, recall, and F1 score. An ablation experiment was carried out to test the contributions of force, image, and position to the recognition performance. The experimental results show that SVM has better performance than the neural network models. However, when the size of the dataset is limited, SVM still faces generalization issues. By applying heuristic action, this paper proposes a two-stage recognition strategy that can improve the recognition success rate of the SVM.

Motion Planning for Cutting Flexible Objects Based on Contact State Recognition

Motion Planning for Cutting Flexible Objects Based on Contact State Recognition

SF6 High-Voltage Circuit Breaker Contact Status Detection at Different Currents.

Currently, the online non-destructive testing (NDT) methods to measure the contact states of high-voltage circuit breakers (HVCBs) with SF6 gas as a quenching medium are lacking. This paper aims to put forward a novel method to detect the contact state of an HVCB based on the vibrational signal. First, for a 40.5-kV SF6 HVCB prototype, a mechanical vibration detection system along with a high-current generator to provide the test current is designed. Given this, vibration test experiments are carried out, and the vibration signal data under various currents and corresponding contact states are obtained. Afterward, a feature extraction method based on the frequency is designed. The state of the HVCB contacts is then determined using optimized deep neural networks (DNNs) along with the method of adaptive moment estimation (Adam) on the obtained experimental data. Finally, the hyperparameters for the DNNs are tuned using the Bayesian optimization (BO) technique, and a global HVCB contact state recognition model at various currents is proposed. The obtained results clearly depict that the proposed recognition model can accurately identify five various contact states of HVCBs for the currents between 1000 A and 3500 A, and the recognition accuracy rate is above 96%. The designed experimental and theoretical analysis in our study will provide the references for future monitoring and diagnosis of faults in HVCBs.

Robot assembly theory and simulation of circular-rectangular compound peg-in-hole

AbstractIn this paper, the assembly problem of circular-rectangular compound peg and hole parts in automatic assembly is proposed for the first time, and an automatic assembly method based on six-dimensional force sensor is summarized. Firstly, according to the quasi-static equilibrium condition, the contact states are summarized. It is concluded that there are 7 categories of 44 contact states in insertion stage and the forces conditions to maintain each contact state; Secondly, according to the analysis of the force/moment relationship, the jamming diagram is drawn, and the conditions of jamming are analyzed. Thirdly, the assembly strategy is discussed, including contact state recognition strategy, hole searching strategy, and pose adjustment strategy. Finally, combined with impedance control, the assembly theory is verified in ADAMS, and the assembly with minimum clearance of 0.1 mm is achieved.

Evaluation of the applica bility of the gaussian distribution mixture model for identification of the contact state in the problems of robotic assembly of "shaft—sleeve" connections

The problem of evaluating the effectiveness of contact state recognition (CS) is solved based on the application of the Gaussian Mixture Model (GMM) in a force-controlled robotic assembly. Using the proposed CS modeling strategy of the GMM model based on the similarity measure of the distribution, a high degree of classifi cation success was obtained. An experimental setup with an ABB IRB-140 industrial robot was used to evaluate the operability of the proposed method.

Fuzzy Logic-Driven Variable Time-Scale Prediction-Based Reinforcement Learning for Robotic Multiple Peg-in-Hole Assembly

Reinforcement learning (RL) has been increasingly used for single peg-in-hole assembly, where assembly skill is learned through interaction with the assembly environment in a manner similar to skills employed by human beings. However, the existing RL algorithms are difficult to apply to the multiple peg-in-hole assembly because the much more complicated assembly environment requires sufficient exploration, resulting in a long training time and less data efficiency. To this end, this article focuses on how to predict the assembly environment and how to use the predicted environment in assembly action control to improve the data efficiency of the RL algorithm. Specifically, first, the assembly environment is exactly predicted by a variable time-scale prediction (VTSP) defined as general value functions (GVFs), reducing the unnecessary exploration. Second, we propose a fuzzy logic-driven variable time-scale prediction-based reinforcement learning (FLDVTSP-RL) for assembly action control to improve the efficiency of the RL algorithm, in which the predicted environment is mapped to the impedance parameter in the proposed impedance action space by a fuzzy logic system (FLS) as the action baseline. To demonstrate the effectiveness of VTSP and the data efficiency of the FLDVTSP-RL methods, a dual peg-in-hole assembly experiment is set up; the results show that FLDVTSP-deep Q-learning (DQN) decreases the assembly time about 44% compared with DQN and FLDVTSP-deep deterministic policy gradient (DDPG) decreases the assembly time about 24% compared with DDPG. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —The complicated assembly environment of the multiple peg-in-hole assembly results in a contact state that cannot be recognized exactly from the force sensor. Therefore, contact-model-based methods that require tuning of the control parameters based on the contact state recognition cannot be applied directly in this complicated environment. Recently, reinforcement learning (RL) methods without contact state recognition have recently attracted scientific interest. However, the existing RL methods still rely on numerous explorations and a long training time, which cannot be directly applied to real-world tasks. This article takes inspiration from the manner in which human beings can learn assembly skills with a few trials, which relies on the variable time-scale predictions (VTSPs) of the environment and the optimized assembly action control strategy. Our proposed fuzzy logic-driven variable time-scale prediction-based reinforcement learning (FLDVTSP-RL) can be implemented in two steps. First, the assembly environment is predicted by the VTSP defined as general value functions (GVFs). Second, assembly action control is realized in an impedance action space with a baseline defined by the impedance parameter mapped from the predicted environment by the fuzzy logic system (FLS). Finally, a dual peg-in-hole assembly experiment is conducted; compared with deep Q-learning (DQN), FLDVTSP-DQN can decrease the assembly time about 44%; compared with deep deterministic policy gradient (DDPG), FLDVTSP-DDPG can decrease the assembly time about 24%.

Symbolic-Based Recognition of Contact States for Learning Assembly Skills.

Imitation learning is gaining more attention because it enables robots to learn skills from human demonstrations. One of the major industrial activities that can benefit from imitation learning is the learning of new assembly processes. An essential characteristic of an assembly skill is its different contact states (CS). They determine how to adjust movements in order to perform the assembly task successfully. Humans can recognize CSs through haptic feedback. They execute complex assembly tasks accordingly. Hence, CSs are generally recognized using force and torque information. This process is not straightforward due to the variations in assembly tasks, signal noise and ambiguity in interpreting force/torque (F/T) information. In this research, an investigation has been conducted to recognize the CSs during an assembly process with a geometrical variation on the mating parts. The F/T data collected from several human trials were pre-processed, segmented and represented as symbols. Those symbols were used to train a probabilistic model. Then, the trained model was validated using unseen datasets. The primary goal of the proposed approach aims to improve recognition accuracy and reduce the computational effort by employing symbolic and probabilistic approaches. The model successfully recognized CS based only on force information. This shows that such models can assist in imitation learning.

Fuzzy inference mechanism for recognition of contact states in intelligent robotic assembly

This paper presents a methodology for generating a fuzzy inference mechanism (FIM) for recognizing contact states within robotic part mating using active compliant motion. In the part mating process, significant uncertainties are inherently present. As a result it is pertinent that contact states recognition systems operating in such environment be able to make decisions on the contact state currently present in the process, based on data full of uncertainties and imprecision. In such conditions, implementation of fuzzy logic and interval inference brings significant robustness to the system. As a starting point for FIM generation, we use a quasi-static model of the mating force between objects. By applying Discrete Wavelet Transform to the signal generated using this model, we extract qualitative and representative features for classification into contact states. Thus, the obtained patterns are optimally classified using support vector machines (SVM). We exploit the equivalence of SVM and Takagi---Sugeno fuzzy rules based systems for generation of FIM for classification into contact states. In this way, crisp granulation of the feature space obtained using SVM is replaced by optimal fuzzy granulation and robustness of the recognition system is significantly increased. The information machine for contact states recognition that is designed using the given methodology simultaneously uses the advantages of creation of machine based on the process model and the advantages of application of FIM. Unlike the common methods, our approach for creating a knowledge base for the inference machine is neither heuristic, intuitive nor empirical. The proposed methodology was elaborated and experimentally tested using an example of a cylindrical peg in hole as a typical benchmark test.

Development of an Arrayed Tactile Sensor Having Four Stories and Recognition of Contact State Using Neural Networks

In the human skin, four kinds of tactile receptors are distributed three dimensionally, i.e., not only on the skin surface but also in the four corresponding depths of the skin. Imitating this, a tactile sensor having four stories is proposed, inside each story of which capacitive force sensing elements are distributed in an array. The recognition method of contact state using neural networks is proposed. The stress distribution of a sensor sheet is obtained, and the outputs of the sensing elements are simulated by FEM. By applying the outputs to neural networks, it is confirmed that the contact shape recognition is possible, and the sensor having four stories achieves higher recognition rate compared with that having one story. The MEMS fabrication method of an arrayed capacitive tactile sensor made of PDMS is proposed, in which Parylene is used as an anti-stiction layer for the purpose of peeling off the PDMS sheet easily from the substrate. By stacking this sensor one by one, fabrication of a sensor having four stories is possible. A sensor having one story is preliminarily fabricated. The basic performance of one force sensing element and that of an arrayed sensor composed of 3×3 elements are characterized, which shows the potential of the proposed sensor for tactile sensing.

2A1-E29 教師あり学習を用いた四層分布型触覚センサによる接触状態の判定

2A1-E29 教師あり学習を用いた四層分布型触覚センサによる接触状態の判定

2P1-S-066 ニューラルネットワークを用いた分布型触覚センサによる接触状態の判定(進化・学習とロボティクス5,生活を支援するロボメカ技術のメガインテグレーション)

2P1-S-066 ニューラルネットワークを用いた分布型触覚センサによる接触状態の判定(進化・学習とロボティクス5,生活を支援するロボメカ技術のメガインテグレーション)

A hidden Markov model-based assembly contact recognition system

A hidden Markov model (HMM)-based assembly contact state recognition system is designed and implemented. The system utilizes the force/torque data captured from a wrist force sensor to extract the intrinsic spatial relationships of contact formations arise from robotic assembly. The paper introduces a theoretical framework for contact recognition based on discrete HMM with special emphasis on the practical realization of the system using an industrial robot. With the detailed exposition of the major algorithms for solving the three key problems in HMM, namely, the evaluation of observation sequence probability, optimization of state sequence probability, and optimization of model parameters, a working prototype of HMM-based system is developed. The performance of the contact recognition system is investigated by using experimental studies. The results obtained clearly demonstrate that HMM is an effective means for assembly contact modelling and identification, and that the framework reported in this paper provides an essential ground work for the development of a practical intelligent robotic system.

接触状態変化検出機能を有するマニピュレーションシステムに関する検討

In a manipulation system that executes assembly tasks, contact-state recognition or contact-state-transition detection are important functions. These functions should be modularized as a motion monitoring and evaluation (MME) module. This paper mainly presents a real-time detection method of contact-state-transition with high reliability, and the implementation of the method into the MME module. In the detection algorithm, features which fluctuate abruptly when the contact-state-transition occur are selected according to the motion pattern. In the detection algorithm, the differential of the feature locus is monitored and the transition region is distinguished from the stable region. Then, to confirm whether state-transition really occur in the region, the duration of the region and the feature deviation are counted. Using this method, contact-state-transition is detected with only 40 [msec] delay. An experimental manipulation system, in which motion control and MME modules are independently executed, is developed using transputers and the 6 D.O.F direct drive manipulator ETA 3. In the experiment, six patterns of motion, in which different types of transition occur, are selected. The experimental results show that contact-state-transition is accurately detected and motion commands can be quickly switched using detected transition information.

Identification of contact states based on a geometric model for manipulative operations

_A model-based approach to the recognition of contact states is presented. The state of contacts between workpieces varies as the operation proceeds in manipulative tasks such as part-mating and grasping. Adequate control laws and motion strategies of a robot manipulator depend on the contact state. The recognition of contact states is thus required so that the robot can select and modify its motion strategies adequately according to the state. In this paper, we deal with the estimation of contact states by using force information acquired in the mating process, and develop a method for generating the state classifiers based on geometric models of workpieces on a computer. First, a symbolic representation of contact states is addressed. Second, the static behaviour of workpieces at each contact state is analysed by applying the theory of polyhedral convex cones. Next, state classifiers that discriminate contact states are formulated by using the polyhedral convex cones, which directly provide a set of dis...