Why, How, What (2.0)

IMG_0195When people describe their job, they often start with the “what”. “What do you do?” is in fact the most likely question you’d formulate to ask about some aspect of your interlocutor’s life. (What do you do is also the first question I learn to formulate in English, I did not progress too much with my vocabulary since then…).

I would start my blog by answering to a more inspiring question instead: why do you do (your job)? I have been spending the last five years looking for an answer. I concluded as follows:

  • I have been always curious about the way natural phenomena happen: a waterfall, the way trees grow, for example.
  • I care about the landscapes of my island.
  • I hate the effect on my territory of some production plants: pollution, abuse of resources… other “collateral” damages.
  • Some doors were open and others weren’t.

My job – advanced process control consultant – is hence the answer to a different question: how do you pretend to solve the problem of a sustainable development of the society, while you curiosity for natural phenomena is satisfied?

And only then the “what” comes.

With this blog I’d like to share my “what”: share interesting anedoct about control room late shift experiences; tell something about my past research activities; review interesting papers; keep you posted about my activities: presentations, conferences… But, don’t worry, there will be space left for more “why”, I am sure!

Enjoy the reading!

(Year 2020)

Quotes

« I take the positivist viewpoint that a physical theory is just a mathematical model and that it is meaningless to ask whether it corresponds to reality. All that one can ask is that its predictions should be in agreement with observation. »

Stephen Hawking and Roger Penrose

Publication Backstage: A distillate composition estimator for an industrial multicomponent IC4-NC4 splitter with experimental temperature measurements

M. Porru, J. Alvarez, R. Baratti, IFAC Proceedings volumes, 46(32), 391-396, 2013

In this publication I discuss the theoretical work underlying the development of the soft-sensor implemented at the oil-refinery Sarlux (Sarroch, Italy) for monitoring the separation of an industrial hepta-component distillation column, which is important because it serves for producing the feed to an alkylation reactor for high-octane gasoline production. The application presented in this paper is an example of a successful applied research in the field of process control. The achievements were possible thanks to a perfect combination between (i) theoretical knowledge on estimability theory, and (ii) communication with industry in phase of design and testing with industrial data. The online performance of the soft sensor after the commissioning are available in my PhD thesis. The soft-sensor is currently returning good estimation performance, after 3 years of operation.

It is also worth mentioning here that model-based estimators for multicomponent distillation columns are generally considered difficult to implement due to the high-dimensionality of the model itself. In this paper I also show how to easily overcome the problem by reducing the model order based on thermodynamics arguments.
I consider the design and implementation of this soft sensor the greatest achievement from the perspective of my applied oriented research (always working towards even better achievements!).

Abstract: The problem of on-line estimating on the basis of temperature measurements the distillate NC4 impurity in an industrial IC4-NC4 splitter is addressed within an adjustable-structure Geometric Estimation approach, yielding:(i) suggestive sensor location guidelines drawn from detectability measures, and (ii) conclusive results obtained from estimator functioning assessment with simulated and experimental data. The resulting estimator performs the estimation task within an admissible error tolerance, and considerably less ODEs than the ones of the standard EKF technique employed in the majority of related previous studies.

Curious? Read more here:
https://www.sciencedirect.com/science/article/pii/S1474667015382896

DISCOVER IMPROVISE: Improved Process Operation via Rigorous Simulation Models

IMPROVISE-Discover.jpg

The IMPROVISE project, that gave me the possibility to move to TU/Eindhoven within the Control Systems Group, is going to end. The highlights of the project have been summarised in the poster “DISCOVER-IMPROVISE”, that I designed with the contribution of my team mates. Thank you everybody for the great experience!

Observability – what is it?

I have been doing this exercise for long time, and I am never satisfied by the result. Now, I try again. What is observability? As process controllers, we all have in mind the definitions of the books. For example we, process controllers, can easily recall, by heart, the definition for linear systems. Given a dynamic model M describing a process (how the hell do I put equations in this blog?)

IMG_0120.jpg

 

 

(being x the vector state of dimension N, u the input, and y the vector of the measurements with dimension m, for simplicity m=1), the system is observable with the model M and the measurement y if the observability matrix O, dim(O)=NxN

IMG_0121.jpg

is full rank (i.e. rank(O)=N). And generally, the story ends with the students calculating O, then its rank. If this rank is full, then they may implement the kalman filter borrowing some Matlab routine. For nonlinear  systems, the story is similar but, instead of having the matrices A, B, C, we have nonlinear maps, which make us invoke the Lie derivatives for the calculation of O. And again, if O is full rank, ok, observability holds.

But then again, what is observability? What does it tell? Well, I would say that observability is a property of the model that, if satisfied, allows the available process measurements together with the model, to fully reconstruct the state of the system.  I try to make the concept more clear making an example. Let me take the non isothermal CST reactor with first order rate for the reaction A–>B (Fig. 1)

IMG_0129

Fig. 1 The CSTR

described by the model equations:

IMG_0128

You may remember that its plane phase, in case of multiple equilibrium point (E1, E2, E3) looks a little bit like Fig. 2.

img_0231

Fig. 2 (Not very accurate-) Plane phase of the CSTR with multiple steady states E1, E2, E3.

Let me assume that the system is travelling in a trajectory that, from the initial condition x0=[T0, C0], is going towards the equilibrium point E1 (Fig 3),

IMG_0235

Fig. 3 The system is approaching the stable steady state E1 following this state trajectory

or in other words: concentration and temperature in the tank are varying until the stable steady state E1 is reached. At the time t, the state of the system is described by the coordinates x(t)=[T(t),C(t)] (Fig. 4).

IMG_0236

The state of the system at the time t

If only temperature measurements are available, the output map is y=T.  In the plain phase y=T is a straight line (Fig. 5). This measurement alone gives only partial information about the state. Infact, it does not say anything about concentration.

IMG_0241

Fig. 5 In the plane phase, y=T is an horizontal line.

However, if we take the first Lie derivative (the CSTR model is nonlinear) of the output map, we can draw a second curve in the phase plane ( ẏ = ∂h/x= f(T,C) ) (Fig. 6). One can note that, in order to draw this curve, the model equation fis required.

IMG_0242

Fig. 6 Another curve can be drawn in the phase plane by using the information given by f(T,C) 

 

 

 

 

 

 

 

 

 

If observability holds, the information carried by this second curve ẏ =  f(T,C)  significantly differs from the one deriving from the first straight line y=T (i.e. ẏ =  f(T,C)  is not coincident nor parallel to the straight line). If observability holds, the intersection between the two curves (Fig. 7) represents the state of the system at the time t (Fig.8):

 

img_0243.png

Fig. 7 The intersection between  y=T  and ẏ =  f(T,C)

IMG_0244

Fig, 8 The intersection between  y=T  and ẏ =  f(T,C) and the state trajectory

 

 

 

 

 

 

 

 

For a two dimensional system (like the considered CSTR), two different curves are necessary to identify the state of the system. This is why the 2×2 observability matrix O is required to have rank 2. If O was rank deficient, it would mean that the second curve is parallel or coincident to the first one, and the intersection point (i.e. the state of the system) does not exist. In other words the system is not observable.

I know that there are more rigorous definitions of observability, but they are too abstracted for me. I much more like to think about observability in the way I just described. I need to thank prof. Jesus Alvarez, who taught  me the observability concepts this way during my stay in Mexico City in 2013.

Coming next: detectability!

 

 

Quotes

The scientist does not study nature because it is useful; [s]he studies it because [s]he delights in it, and [s]he delights in it because it is beautiful. If nature were not beautiful, it would not be worth knowing, and if nature were not worth knowing, live would not e worth living.

Poincare’

Enlightening paper about food chemistry

As recently involved in food processing within the INSPEC (Integrating Sensor Based Process Monitoring and Advanced Process Control) project, I am doing a lot of literature review to get a better understanding of the phenomenological aspects of the diary industry. I read many papers focused on process modeling. When new ideas started running I realized that I could not do simple thought experiments. Why? because foods do not behave straight, like hydrocarbons… While looking for a good book that describes the phase diagram of lactose (did not find any yet, so maybe you can help here…), I stumbled upon a great publication:

Food structure and functionality: a soft matter perspective

Ubbink, J., Burbidge, A., & Mezzenga, R. (2008). Food structure and functionality: a soft matter perspective. Soft matter4(8), 1569-1581.

It tells about food structures, giving a very nice overview of phenomena acting on different length scales. Particularly interesting for me is the section related to the amorphous state.

Surprising (but it should not have…) is for me the concept of “memory of process history”, meaning that these type of systems cannot be completely described by they chemical composition. Here, it comes a great example: take identical recipes for pizza dough, use different ways of mixing (let say manually or with the mixer), and you will get different baked pizza bases (the paper has pictures of this!). This definitively convinced me that to get mayonnaise I have to pay attention to the trajectory in the process space…(well, next time). Behind my improvements in cooking skills, this has implications is scaling food processes up to factories. The section about food processing concludes with a reflection on what most bother me:

what model should I use to understand my process? how deep do I have to dig? What is the level of simplification I can allow to have a model for control?

Looking forward to developing models for control of diary processes!

Marcella

P.S. I also want to quote the consideration that “only two of the common foods have evolved with the primary purpose of being eaten“: milk and fruit. “All other foods and food ingredients from nature with a complex structure have evolved for reasons having nothing to do with their use as a food”. We have all the ingredients to start a new diet

: ) !!

Why, How, What

IMG_0195When people describe their job, they often start with the “what”. “What do you do?” is in fact the most likely question you’d formulate to ask about some aspect of your interlocutor’s life. (What do you do is also the first question I learn to formulate in English, I did not progress too much with my vocabulary since then…).

I would start my blog by answering to a more inspiring question instead: why do you do (your job)? I have been spending the last five years looking for an answer. I concluded as follows:

  • I have been always curious about the way natural phenomena happen: a waterfall, the way trees grow, for example.
  • I care about the landscapes of my island.
  • I hate the effect on my territory of some production plants: pollution, abuse of resources… other “collateral” damages.
  • Some doors were open and others weren’t.

My job – researcher in the field of process control – is hence the answer to a different question: how do you pretend to solve the problem of a sustainable development of the society, while you curiosity for natural phenomena is satisfied?

And only then the “what” comes.

With this blog I’d like to share my “what”: tell something about my research; review interesting papers; keep you posted about my activities: presentations, conferences… But, don’t worry, there will be space left for more “why”, I am sure!

Enjoy the reading!

(Year 2017)