On the dynamics of human locomotion and co-design of lower limb assistive devices

Recent developments in lower extremities wearable robotic devices for the assistance and rehabilitation of humans suffering from an impairment have led to several successes in the assistance of people who as a result regained a certain form of locomotive capability. Such devices are conventionally designed to be anthropomorphic. They follow the morphology of the human lower limbs. It has been shown previously that non-anthropomorphic designs can lead to increased comfort and better dynamical properties due to the fact that there is more morphological freedom in the design parameters of such a device. At the same time, exploitation of this freedom is not always intuitive and can be difficult to incorporate. In this work we strive towards a methodology aiding in the design of possible non-anthropomorphic structures for the task of human locomotion assistance by means of simulation and optimization. The simulation of such systems requires state of the art rigid body dynamics, contact dynamics and, importantly, closed loop dynamics. Through the course of our work, we first develop a novel, open and freely available, state of the art framework for the modeling and simulation of general coupled dynamical systems and show how such a framework enables the modeling of systems in a novel way. The resultant simulation environment is suitable for the evaluation of structural designs, with a specific focus on locomotion and wearable robots. To enable open-ended co-design of morphology and control, we employ population-based optimization methods to develop a novel Particle Swarm Optimization derivative specifically designed for the simultaneous optimization of solution structures (such as mechanical designs) as well as their continuous parameters. The optimizations that we aim to perform require large numbers of simulations to accommodate them and we develop another open and general framework to aid in large scale, population based optimizations in multi-user environments. Using the developed tools, we first explore the occurrence and underlying principles of natural human gait and apply our findings to the optimization of a bipedal gait of a humanoid robotic platform. Finally, we apply our developed methods to the co-design of a non-anthropomorphic, lower extremities, wearable robot in simulation, leading to an iterative co-design methodology aiding in the exploration of otherwise hard to realize morphological designs

Modeling and Simulation of Energy Efficient Milling Process Plans for Prismatic Parts

Nenad Nesic

The milling process has been continuously optimized from many aspects: forces, velocities, stability, quality. Numerous papers have been published that report results in the domain of toolpath optimization with criteria such as raising the quality of the machined part, obtaining a stable machining process and maximizing the material removal rate. However, only recently researchers have started investigating the ecological impact of machining processes. With the constantly increasing prices of electrical energy and all environmental problems caused by the production and waste of the energy, it has become indispensable to look for ways to optimize machining processes from the energy consumption aspect, too. It is in this domain, then, that the work described in this dissertation contributes – to add a new criterion for the sustainable operation of machine tools: reducing the energy consumption during the use phase of the machine tool. The total energy spent in a machining process is the sum of power spent by all machine tool subsystems multiplied by the time that they are working. Therefore, the operating strategies for increasing energy efficiency for the high-speed machining processes must be oriented towards the reduction of the total machining time (productive and unproductive) and reduction of instantaneous consumed power in the machining process or, preferably, both. There are many causes for milling process time and energy inefficiency. The major sources of potential time reductions include: (i) the issues related to the geometry and topology of the toolpath („air-time”, overlapping segments, gouge and uncut regions, orientation of the toolpath), (ii) the issues related to the kinematics of the feed drives (the feed velocity/acceleration profile). The possible domains for power consumption reduction are related to the forces that exist in the machining process (cutting, inertial, gravitational and friction forces). The power invested to overcome the cutting force is necessary for the machining process but its consumption should be optimized through constant cutting engagement and feedrate scheduling strategies. The other forces are responsible for the pure mechanical power loss and they should be minimized. The optimization variables for inertial and gravitational forces include the feed motion profile and moving mass configuration, while friction losses are load and/or velocity dependent. In addition, there are some power losses in the electric components of machine tool drives (electrical power losses), whose dependency on velocity and mechanical load can also be modeled. Despite many research efforts to analyze and model the phenomena causing inefficiency of the machining process, they are usually not taken into account during milling process planning. In fact, rare are the attempts to use this knowledge to optimize the milling process with more than one of the above mentioned criteria. The reason for that lies in the fact that increasing the instantaneous values of tool engagement and feedrate in order to minimize total machining time also increases the consumed power and, indirectly, deteriorates the tool and/or destabilizes the cutting process. The existence of conflicting criteria in the objective functions imposes the use of evolutionary algorithms for multiobjective optimization. The process of experimental validation of such an optimization model requires costly and complex external hardware setup and it is also time consuming. Hence, there is an emerging need for the development of an accurate model, which would be used for the simulation of the milling process and testing the competing optimization strategies. The ultimate goal of contemporary manufacturing science and engineering is the development of a holistic virtual machine tool system. The motivation of this PhD research is to contribute to this objective by developing the “virtual milling process for prismatic parts (produced by 2.5D milling)”. This category of objects is selected, because the majority of CNC milling tasks can be performed using 2.5D milling. A very large number of mechanical parts are prismatic (their geometry consists exclusively of features that represent 2D contours extruded in a perpendicular direction). Parts that are even more complex are usually produced from a blank by a 2.5D roughing and 3D–5D finishing. Therefore, development of an accurate model of a 2.5D milling process represents a significant contribution to manufacturing engineering. Starting with above given problem description, we have defined the following research objectives: 1. Model power and energy consumption of a 2.5D milling machine as a function of the toolpath geometry (curvature and continuity), cutting engagement, real feedrate profiles and spindle rotational speed, taking into account moving mass configuration, all power losses (mechanical and electrical) in feed-axis and spindle drives and electromechanical constraints of servo drives; 2. Design an integrated software environment for simulation of the developed models; 3. Validate the model experimentally; 4. Demonstrate the potential of its use for multi-objective process optimization. Cutting engagement calculation is performed by using two competing methodologies, in order to make a comparison of their performances. The first one, the pixel based method, is based on discrete image processing, while the other, the exact method, is based on Boolean operations and vector geometry. The tests performed in more than 20 feature/toolpath combinations demonstrate that the exact method is more precise whereas the pixel-based method is computationally much faster. The machine tool operator commands a constant feedrate value. During the process, based on his/her subjective judgment, he/she can manually tune it to keep the process in the desired boundaries. One of the objectives of this research is to predict the real feedrate values so that they can be integrated in the part program before the process start. The feasible profiles of feed velocities for each toolpath segment are analyzed by answering the following questions: can the commanded feedrate be achieved and with what sequence of acceleration and deceleration phases, how much the tool has to slow down in corners and how does the feedrate profile degenerate in segments where the commanded feedrate cannot be reached? In order to determine all forces and torques, a machine tool model is developed. The following subsystems are modeled as rigid bodies (without vibrations): feed motor shaft, transmission, lead screw/nut pair, lead screw bearings, guides (sliding and rolling design), table with the workpiece and electrospindle (rotor, stator and bearings). All kinetic variables (linear and angular speeds; cutting, inertial and friction forces and torques) are then calculated in the joints of this multi-body system. The modeling of all speeds, forces and torques that occur in the machine tool system during the machining process, allows for calculation of various indicators of process time and energy efficiency: material removal rate, total machining time, useful power (invested in cutting), power loss due to the friction in the joints, power losses in electrical motors, electrical energy consumption of feed and spindle drives. In order to assess the accuracy and the reliability of the developed model two sets of experiments have been performed. The model validation is performed on several subsystem levels and the simulation results show a very good compliance with the experimental results. The first experimental setup consists of the force dynamometer, to measure cutting force components in linear axis directions, two laser optical position sensors, to measure the real feedrate and the power sensor, to measure the total power consumption of the machine tool electrical motors. The machining process was performed on the 6-axis machining center C.B. Ferrari A152 and the data acquisition platform was developed in LabVIEW 2010. The second experimental setup consists of the torque dynamometer, to measure the cutting torque directly on the spindle, the built-in current sensor that measures the motor current proportional to the electrical power consumed by the spindle and, again, the power sensor, to measure the total power consumption of the machine tool electrical motors. This time, MIKRON HPM 600U, a 5-axis milling machine equipped with the controller iTNC530, was used for the experimentation. The developed model is ready for practical implementation. Several examples have been prepared to demonstrate its capabilities in the domains of milling process simulation, sensitivity analysis and prospective optimization strategies. These examples include the prediction of various kinetostatic variables for given toolpaths, choice of the optimum machining strategy (minimization of machining time, moving mass and energy consumption) and process optimization by feedrate scheduling (reduction of cutting force fluctuation). As a conclusion, the major contribution of this PhD research is the development of a comprehensive rigid-body dynamics model of a machine tool system, aimed for the generation, simulation and optimization of 21⁄2D milling process plans. The model features several novelties compared to existing approaches: (i) it takes into account the machine tool rigid body dynamics and real tool kinematics in toolpath planning, (ii) it predicts all mechanical and electrical power losses in the machine tool system (iii) it allows the modeling of machine tool system electromechanical constraints. The model was experimentally validated on two real machine tools in different workshops and its capabilities for the simulation of different toolpath patterns and optimization of process parameters were successfully demonstrated using several examples.

EPFL2012

Pattern generators with sensory feedback for the control of quadruped locomotion

Auke Ijspeert, Ludovic Righetti

Central Pattern Generators (CPGs) are becoming a popular model for the control of locomotion of legged robots. Biological CPGs are neural networks responsible for the generation of rhythmic movements, especially locomotion. In robotics, a systematic way of designing such CPGs as artificial neural networks or systems of coupled oscillators with sensory feedback inclusion is still missing. In this contribution, we present a way of designing CPGs with coupled oscillators in which we can independently control the ascending and descending phases of the oscillations (i.e. the swing and stance phases of the limbs). Using insights from dynamical system theory, we construct generic networks of oscillators able to generate several gaits under simple parameter changes. Then we introduce a systematic way of adding sensory feedback from touch sensors in the CPG such that the controller is strongly coupled with the mechanical system it controls. Finally we control three different simulated robots (iCub, Aibo and Ghostdog) using the same controller to show the effectiveness of the approach. Our simulations prove the importance of independent control of swing and stance duration. The strong mutual coupling between the CPG and the robot allows for more robust locomotion, even under non precise parameters and non-flat environment

2008

Rich and Robust Bio-Inspired Locomotion Control for Humanoid Robots

Nicolas Benoît Dominique Van der Noot

Bipedal locomotion is a challenging task in the sense that it requires to maintain dynamic balance while steering the gait in potentially complex environments. Yet, humans usually manage to move without any apparent difficulty, even on rough terrains. This requires a complex control scheme which is far from being understood. In this thesis, we take inspiration from the impressive human walking capabilities to design neuromuscular controllers for humanoid robots. More precisely, we control the robot motors to reproduce the action of virtual muscles commanded by stimulations (i.e. neural signals), similarly to what is done during human locomotion. Because the human neural circuitry commanding these muscles is not completely known, we make hypotheses about this control scheme to simplify it and progressively refine the corresponding rules. This thesis thus aims at developing new walking algorithms for humanoid robots in order to obtain fast, human-like and energetically efficient gaits. In particular, gait robustness and richness are two key aspects of this work. In other words, the gaits developed in the thesis can be steered by an external operator, while being resistant to external perturbations. This is mainly tested during blind walking experiments on COMAN, a 95 cm tall humanoid robot. Yet, the proposed controllers can be adapted to other humanoid robots. In the beginning of this thesis, we adapt and port an existing reflex-based neuromuscular model to the real COMAN platform. When tested in a 2D simulation environment, this model was capable of reproducing stable human-like locomotion. By porting it to real hardware, we show that these neuromuscular controllers are viable solutions to develop new controllers for robotics locomotion. Starting from this reflex-based model, we progressively iterate and transform the stimulation rules to add new features. In particular, gait modulation is obtained with the inclusion of a central pattern generator (CPG), a neural circuit capable of producing rhythmic patterns of neural activity without receiving rhythmic inputs. Using this CPG, the 2D walker controllers are incremented to generate gaits across a range of forward speeds close to the normal human one. By using a similar control method, we also obtain 2D running gaits whose speed can be controlled by a human operator. The walking controllers are later extended to 3D scenarios (i.e. no motion constraint) with the capability to adapt both the forward speed and the heading direction (including steering curvature). In parallel, we also develop a method to automatically learn stimulation networks for a given task and we study how flexible feet affect the gait in terms of robustness and energy efficiency. In sum, we develop neuromuscular controllers generating human-like gaits with steering capabilities. These controllers recruit three main components: (i) virtual muscles generating torque references at the joint level, (ii) neural signals commanding these muscles with reflexes and CPG signals, and (iii) higher level commands controlling speed and heading. Interestingly, these developments target humanoid robots locomotion but can also be used to better understand human locomotion. In particular, the recruitment of a CPG during human locomotion is still a matter open to debate. This question can thus benefit from the experiments performed in this thesis.