Views
10 months ago

MDA Technologies 4/2014

MDA Technologies 4/2014

Measurement and control

Measurement and control Intelligent assistance system to support the start-up procedure of electro hydraulic drives Ulrich Walter The commissioning of hydraulic controlled axes is often a time-consuming and therefore cost-intensive activity, as particularly a system of this kind covers a number of technologies, and therefore it is not always clear who should be responsible for this technology. Although there is much talk of intelligent hydraulic axes, at the same time the complexity is increased so much that only an expert can understand it. Author: Ulrich Walter ist managing director of the W.E.St. Elektronik GmbH in 41372 Niederkrüchten 1 Introduction For us, intelligence begins at that point where the user can be relieved of complex tasks. With the assistant presented here for analysing and parameterising the hydraulic axis, we take the step of providing the engineer, the mechanic/electrician or the hydraulics engineer with a tool which enables the main characteristic data to be determined and the parameters to be set in a r eliable manner. We are all familiar with assistance systems in modern cars which help the driver to cope with critical situations more easily. So what is a critical situation in a hydraulic system? The commissioning, of course (particularly when axes have not been calculated or simulated in advance). Possible faults, ranging from an incorrectly fitted check valve, pipework or cabling mistakes and incorrect design (unsuitable components for the application) due to the absence of information on the dynamic behaviour of the system, often only enable commissioning to be carried out in an experimental manner, which is then very time consuming. However, an intelligent electronic system can experiment more quickly, better and always using the same logical concept. The software module for automatic commissioning analyses the system behaviour and makes the appropriate decisions, which then lead to reliable system performance. This has less to do with the mathematical analysis of the system (this is only necessary when setting up the control parameters), but rather the logical steps involved in commissioning and working out what to do when something unexpected occurs. No additional sensors are required to implement all this within an acceptable cost frame; the system only evaluates information from the position sensor. The polarity, speeds and characteristic curves, the zero drift, any positive overlap and the

Measurement and control system dynamics are determined during the optimisation process. System parameters and control system gains are adjusted based on these data. At the same time, this assistant makes the information that has been determined available to the user. With a functioning system, this can be done without any problems, but if the assistant has to terminate the optimisation process prematurely, then the reason for the premature termination must be analysed as accurately as possible in order to rectify the fault. When implementing the commissioning assistant, it has been established that it is not the quality of the dynamic optimisation, but rather the profane points which the human being sees at a glance which are a measure of the intelligence; however these have to be laboriously instilled into a software package. In this paper, we will describe the development of this assistant and present practical results. 2 The problem The user, the system designer or the application engineer has designed a hydraulic positioning drive (as is often the case based on empirical values, available components and good feeling). This activity is not trivial, as knowledge about the hydraulics, the measuring equipment, the electronics and the machine and the application are required. This is not part of the day-to-day business, particularly in small companies. The situation then arises during commissioning where the drive does not work satisfactorily and no one really knows „why“. It may be due to workmanship, design faults or only a problem of understanding in the use of the different components. As the system is a dynamic one, many points are difficult to evaluate and must be analysed dynamically. The assistant cannot solve design problems. But the assistant is able to give advises about solving the problem. 2.1 Assumptions The conditions for the use of such tools are acceptable costs and a defined operational area. If the requirements are too high, such developments will never be completed or never be used. If the cost for the customer is too high, the tool will not be accepted. The intention is not to achieve a 100 % solution, but a realisable solution which is suitable for everyday use. The objective is - for the assistant - to be able to work as a modular software tool, even on inexpensive microcontrollers. 2.1.1 On-line versus off-line optimisation? With online optimisation, the system searches continuously for the optimum settings. Theoretically, this would be an ideal system if the residual risks did not have to be taken into account. These risks and the advantages are specified as follows: n Variable masses lead to an optimum setting in each case. Advantage: With small masses, the axis can be moved with higher control dynamics and therefore more quickly. Disadvantage: The loop gain is variable and therefore the system behaviour is difficult to predict. n The dynamic behaviour of the hydraulic drive changes depending on the working point. Tracking errors, control gain and acceleration times are variable. Advantage: Theoretically optimum settings depending on the working point. Disadvantage: A dynamically variable setup which cannot be simple investigated if problems occur. n Interference signals must be detected, as otherwise the optimisation process returns incorrect results. n Reproducible behaviour of the drive with dynamically varying control parameterisation can only be ensured in conjunction with/in a complete motion control system. n A high power of the microcontroller for the online optimisation must be available. It is very difficult to design a complete online optimisation system in such a way that it is capable of working in conjunction with other sub-systems without any problems. However, individual items, such as the speed gain for example, can be adapted online (see MR controller/5/). The approach to the solution presented here is intended primarily to assist commissioning and to detect and solve (if possible) the problems which can occur during commissioning. The commissioning assistant is designed as an off-line solution. 2.1.2 The basic requirements Unfortunately, there is an inversely proportional relationship between the required information, which is provided to the assistant, and the costs. Our approach assumes that there should be no additional costs for sensors and measuring equipment. This results in the following requirements: n No further sensors are used. n A stable control behaviour is to be achieved and not the best possible dynamic setting. n Optimisation is carried out in a defined working range which is limited by the position 1 and position 2. n Optimisation is carried out in open-loop state. n If the optimisation has to be terminated prematurely, information which is as accurate as possible is to be generated for the user. n The user should only has to parameterise the sensor, define the stroke, and set the required speed demand (if available) as a reference. 2.1.3 What are the limits? It is very difficult to assess the limits, as investigations have only been carried out on example systems and by means of HIL. However, two generally critical points have been worked out. n Very slow systems do not usually provide an adequate speed signal resolution to enable the behaviour to be evaluated correctly. These systems are also often very uncritical and meet the requirements without challenging parameterisation. n Extremely fast systems (e.g. stroke times for 100 % stroke of 50 to 200 ms) often do not leave the acceleration/deceleration phase. The method presented is also unsuitable for these systems. 2.2 What can be optimised automatically? Ideally, the points which are difficult to measure are optimised automatically. A control loop with reverse polarity is difficult to detect, for example, if the drive is already at the end of its stroke. The drive remains in this position and does not move. If fault finding is started at a wrong point (after all, there are a large number of possibilities), this can waste a lot of time before the correct approach is found. An MDA Technologies 4/2014 51

E-PAPER KIOSK:

WORLD OF INDUSTRIES - MOTION, DRIVE & AUTOMATION 1/2017
WORLD OF INDUSTRIES - MOTION, DRIVE & AUTOMATION 2/2017
WORLD OF INDUSTRIES - MOTION, DRIVE & AUTOMATION 3/2017
WORLD OF INDUSTRIES - MOTION, DRIVE & AUTOMATION 4/2017
WORLD OF INDUSTRIES - MOTION, DRIVE & AUTOMATION 5/2017
Automation Technologies 1/2016
Automation Technologies 2/2014
Automation Technologies 2/2015
Automation Technologies 2/2016
Automation Technologies 3/2014
Automation Technologies 3/2015
Automation Technologies 3/2016
Automation Technologies 4/2014
Automation Technologies 4/2015
Automation Technologies 4/2016
Automation Technologies 5/2014
Automation Technologies 5/2015
Automation Technologies 5/2016
WORLD OF INDUSTRIES - Industrial Automation 1/2017
WORLD OF INDUSTRIES - Industrial Automation 2/2017
WORLD OF INDUSTRIES - Industrial Automation 3/2017
WORLD OF INDUSTRIES - Industrial Automation 4/2017
WORLD OF INDUSTRIES - Industrial Automation 5/2017