Thrust 3 - Effectiveness
3B.3: Active Vibration Damping of Mobile Hydraulic Machines
Prof. Andrea Vacca, Purdue University
Statement of Project Goals
The goal of the project is to develop a novel energy-efficient control methodology to reduce vibrations in hydraulic machines. The proposed control strategy has potential to replace or limit costly and energy dissipative methods currently utilized to achieve acceptable dynamical behavior in mobile fluid power (FP) applications. The novel solution would allow for a reduction of both amplitude and duration of actuators oscillation up to 70%. The solution offer also margin of energy consumption reduction. Based on an adaptive control method based on pressure feedback (using pressure sensors located in well protected locations of the machine), the proposed techniques is suitable to all mobile applications without introducing significant cost increase.
Project's Role in Support of the Strategic Plan
This project is related to the “control and controllability” topic, and addresses the major technical barriers of “control and energy management” and “efficient systems”. In particular, the project proposes an innovative adaptive electro-hydraulic (EH) control methodology for general application to fluid power machines that permits to reduce machine vibrations according to a general and inexpensive technique that addresses the inherent nonlinearities of the hydraulic systems and the unpredictable operating conditions of the machine (e.g. varying inertia of the load, terrain roughness, variable geometrical configuration of the booms, etc.).
The proposed control method has positive implications as concerns safety, efficiency, controllability and productivity of current FP machines. Moreover, the novel EH method will allow: a) the simplification of current hydraulic circuits, through the removal of elements normally introduced to improve system dynamical behavior; b) to enlarge the area in which fluid power technology can be conveniently applied.
Description and Explanation of Research Approach
FP systems are routinely affected by oscillatory dynamics of moving parts which can lead to stability issues. Undesired vibrations not only worsen controllability, but also reduce productivity and impact comfort and safety of operation. Despite the research efforts in finding solutions to control such oscillations, a general solution has not been found yet. Current damping methods are designed for specific applications and they can damp oscillations only in a limited range of operating conditions. Additionally, they usually introduce systems slowdown (capacitive methods) or energy dissipation (resistive methods). A first distinction among methods for vibration dampening is whether the hydraulic system is integrated with electronic control logic. According to this classification, there are pure hydraulic (PH) solutions and electro-hydraulic (EH) solutions.
Pure hydraulic technology: these methods are based on capacitive elements (e.g. accumulators) and/or resistive elements (e.g. orifices). Being based on fixed parameters these methods are suitable to damp the system only within a small range of operating conditions, and the tuning is typically based on extensive “trial and error” empirical processes made for each single application. The literature reports numerous methods that belong to the pure hydraulic technology category, including the use of accumulators or restrictors. Of particular interest for this research are the dissipative methods based on the applications of counterbalance valves [1, 2].
Electro-Hydraulic (EH) technology: EH technology is based on an optimal management of the power source with respect to a feedback signal representative of the oscillation extent. EH technology has often been combined with PH technology, to extend the range of stability of the hydraulic system and/or to limit the contribution of pure-hydraulic techniques drawbacks on the entire system. Examples are: active suspensions; earthquake simulators, vehicles braking systems; hydraulic robots; active damping seats.
This research particularly investigates the pressure feedback control methodology, in which the pressure signal is used to indirectly quantify the oscillation. In this case a control based on real-time identification of the relationship between pressure and oscillations is required. Some interesting results obtained in the past have not reached practical application because of the complexity of the proposed controllers and of its model-based nature, which makes it difficult to extend to other applications.
The solution of the drawbacks of the past proposed pressure-feedback techniques represent the main challenge of this project. In particular, the proposed control methodology will address the problem of oscillation damping of FP machines considering:
- The uncertainties typical of FP machines (unpredictable load mass, machine varying kinematics, terrain roughness, etc.) and inherent nonlinearities of the hydraulic actuation systems. For this reason the control methodology will be adaptive and not model based;
- The need for formulating a control method that ensures stability and performance over the entire range of operating conditions. This is the crucial limit of current adaptive solutions in FP applications. For this reason the adaptive control will be based on Extremum Seeking control methodology in an innovative way in the FP field;
- Functionality, reliability and cost requirements of FP applications. For this reason the proposed control methodology will be formulated for pressure feedback control (pressure sensors used as feedback signal), overcoming the limits of current position tracking control methods for harsh applications.
The proposed control solution and its innovative contents
This research applies for the first time to FP applications the adaptation/optimization scheme using the Extremum Seeking (ES) theory. ES is an algorithm able to identify the set of parameters that can produce the maximum or minimum of a given function. Figure 1 and Figure 2 describe the idea under the proposed control approach. A controller is used to control the input signals of the control elements (flow control valves, for the case considered in this research), Figure 1. The input parameters of the controller are signals given by pressure sensors installed near the actuator for which the oscillations have to be minimized. The tuning of the control parameters is achieved through online or offline optimization methods. Figure 2 represents the idea for the optimization according to the offline scheme. A cost function associated to the oscillation is evaluated using real experiments or computer simulations. The ES algorithm is used to achieve minimum oscillation through a fast convergence loop.
Figure 1: Basic schematic of the controller
Figure 2: Controller tuning, the optimization process
The project will focus on the investigation of both the real-time and on-line application of the controller to mobile machine. A successfully online implementation will permit to apply the control technique to a wide range of possible real-world machines.
Two important aspects of the project are to investigate the stability and observability of the control strategy related also to the safe performances of the entire system.
The study of hardware requirements (EH valve bandwidth, speed of electronic controller that implements the scheme of Fig. 1), as well as specifics of ES high-pass filter, demodulator, and low pass filter, designed on the basis of the system dynamics, is a particular topic of this research. The aim of the ES optimization scheme is to permit the creation of fast controllers, based on parameters predetermined with ES-based methods. However, this application of ES only partially meets the main goal of the project of obtaining a controller suitable for every application without the necessity of a pre-tuning phase. This especially is the case of the on-line control mode, in which the ES algorithm alone is not able to guarantee the effectiveness of the controller. For this specific topic on-line identification techniques are investigated in order to improve the convergence speed of the algorithm.
The project verifies the performance of the proposed strategy on particular reference cases in which all sources of oscillations can be excited (e.g. support, hanging load, hydraulic circuit): a hydraulic crane (Figure 3).
Figure 3: Instrumented hydraulic crane installed at Maha Lab
Margin of energy consumption reduction and system simplification
An alternative method to classic hydraulic solutions for controlling oscillations can offer significant advantages in terms of energy consumption and system simplification. The hydraulic crane considered in this research is suitable to prove the energetic advantages of the proposed technology. As described in [1, 3] an acceptable dynamic behavior for this kind of machine operating with suspended load is obtained with a setting of the counterbalance valves (in terms of cracking pressure and pilot ratio) that introduces an additional term of energy consumption in addition to the energy dissipation necessary to handle the load (Figure 4). No significant research has been done to quantify this source of energy consumption. After evaluating this contribution, the research will show the benefit in utilizing the new control methodology to simplify the design of counterbalance valves and increase their energetic performance.
Figure 4: Characteristic plot of a counterbalance valve. Losses introduced by its setting.
- Chapple, P. and D. Tilley, Evaluation Techniques For The Selection of Counterbalance Valves, in The Expo and Technical Conference for Electrohydraulic and Electropneumatic Motion Control Technology1994: Anaheim, USA.
- Overdiek, G., Design and Characteristics of Hydraulic Winch Controls by Counterbalance Valves, 1981.
- Boschrexroth. Load Holding - Motion Control RE 00171/02.07. [cited 2012 September]; Available from: http://www.boschrexroth.com.