The PID Tuner is a software tool to help you tune a PID controller optimally, using a step-by-step method. Click HERE to get more information.



We provide consultancy services, combined with implementation and testing of control algorithms or changes to the control algorithm. To get an idea of what we do and what it can bring, please go through some of our projects listed below.  Almost every project description presents (scaled) measurements to show proof of the improvements that we realised.

Customer: ISPT, the Netherlands. ISPT, an abbreviation of the Institute for Sustainable Process Technology, is a so-called top knowledge institute, i.e. an institute where big and small companies, and universities cooperate, and are co-funded by the Dutch government. In 2011 DotX got involved in the ISPT project PH-00-04, titled “Intelligent Observer and Control for Pharmaceutical Batch Crystallization”. The aim of this project (in which 12 companies and knowledge institutes participate), is to develop and test improved batch control. This aim is to be achieved by (1) improving the seeding technique, (2) applying and testing on-line measurement techniques for crystal size distributions and for solution concentration, and (3) applying and testing methods to control crystal growth rate and crystal size. dotx platf meas skid By the end of 2010 a measurement skid had been realised that contains four different on-line measuring sensors and a pump to circulate suspensions of crystals. The skid can be connected to any batch reactor via hoses, see the figure above. DotX has integrated the software and computer systems in a Control platform. This platform allows to perform a wide range of tasks among which reading all measurements on-line, allow on-line trending of any measured variable, and allow researchers to program algorithms of any complexity in their favourite software environment (Matlab), to control the batch sequencing, seeding and coolant during a batch. In essence, the Control platform integrates a multitude of software and hardware products spread over different PCs and PLCs, using three different communication protocols.  
Customer: Tata Steel, the Netherlands. In a Blast Furnace exhaust gases escape during the process of iron making. These gases pass, among others, through a gas turbine to produce electrical power. Ideally, the power output of the turbine is maximised, while the pressure in the Blast Furnace (which is influenced directly by the gas turbine), remains as constant as possible. DotX has modelled the processes, using a fast grey-box modelling method, and, based on the model, proposed several changes to the existing control system. The DotX advices have been implemented by internal Tata staff, and put into operation. Since then, the gas turbine connected to the Blast Furnace produces at 99% of its electrical capacity.
Cokes oven. Customer: Tata Steel, the Netherlands. Cokes furnaces are primarily producing cokes, for ironmaking. In this process, however, gasses are produced as well. It is important that the pressures in the furnaces are kept at target with minimal variations. DotX investigated the cause of the gas pressure variations and the possibilities to reduce these by adjusting the control system. In 2008, the investigations started. It appeared that the control hardware was not suitable for the required changes. 2010, after a replacement of the control hardware, the PI controllers were extended with D-action, and the PID controllers were tuned optimally. PID autotuners were implemented to maintain optimal settings of the resulting PID controllers. Furthermore, Learning Feedforward was added (read more on these solutions this link). As a result, the pressure peak errors were reduced by 50%, and the standard deviations of the gas pressure variations were reduced by 40%. The figure below shows measurements of the pressure in blue (scaled) BEFORE (i.e. without learning FF), and AFTER (with Learning Feedforward). The green line is the trigger signal; if it jumps to 1, the furnace is filled with coal. The Learning Feedforward learns to compensate the resulting pressure variations. Note that all control solutions were implemented in the Tata PLC system. Since 2011, our efforts are more focussed towards the control of other installations in the cokes furnace.  
VDL Weweler is a producer of automotive suspension systems and axles. In 2010, after a feasibility study, DotX was assigned by VDL Weweler to implement DNPC to one of the Weweler walking beam furnaces. In January 2011, the controller was operational. A reduction of at least 4.8 % on average was measured. Further reductions are foreseen, after minor hardware changes. A paper on this subject can be found in the Paper section.This movie shows the HMI (web-interface) of the furnace controller during production at VDL-Weweler while in operation on the VDL-Weweler reheat furnace (no simulation, actual production). Each blue bar respresents a steel product. Each yellow bar shows the temperature of that product. We have increased the playback speed.
Crown van Gelder is a paper producer. When paper is produced from pulp, the grinding of the pulp must be performed with minimal fluctuations, while the pulp level in the mixer (tank) must remain within fixed limits. DotX has changed the control system in January 2008 in such a way that the grinding fluctuations reduced by more than 40%. The figure below shows (scaled) measurements.  
In wastewater treatment plants, accurate control of O2 levels is essential. If the O2 levels vary too much, the bacteria in the plant cannot properly do their work,which is to convert organic material into nitrite and nitrate. Up to 2012, the O2 level was controlled by a standard PID controller. To improve control, we implemented the High Performance PID controller in the PLC. The figure below shows measured results of BEFORE (i.e. with standard PID) and AFTER (with High Performance PID). The latter controller is now in use.
A tailor made controller has been developed for the control of two pumpstations in the polder 'Waterlandse Boezem'. The controller had to satisfy complicated on/off requirements, and had to minimise pumping activity in the more expensive day-hours. A paper on this subject has been published, and can be downloaded in our Paper section. Movie on this subject:The movie shows iterations of the controller while finding the optimal pump settings to maintain the water level (lowest graph) within the limits (dashed lines). The upper graph shows the predicted "load" on the polder (i.e. the expected inflow of water in the next 24 hours).

  1. Modeling of irrigation and drainage canals for controller design, 1999
  2. Simple water level controller for irrigation and drainage canals, 1999
  3. Robust Mould Level Control, ACC 2005
  4. Continuous nonlinear model predictive control of a hybrid water system, 2013
  5. Nonlinear Model Predictive Control of Walking Beam Furnaces, presented at METEC InSteelCon 2011