Optquest for Arena Simulation
Hello friends, greetings to all. In this article, we will first create a simulation model with a single software tool, and then we will perform the optimization process on our model.
Before moving on to the optimization process with the Optquest tool, we should prepare our simulation model. Because the optquest tool will run our simulation model by adhering to the constraints we will create to optimize our objective function and show us the best result.
How about working with a very simple simulation model example? So let’s take a look at our example first.
Example with Optquest Tool
Let’s consider a workshop running on a single assembly line. In this workshop, 3 different processes are carried out and all products produced, generate an income of $ 15. It will be predicted that a part comes to the system every 5 minutes in a fixed manner. The owner wants to optimize his income. Let us consider these processes in detail.
First Process: Two people are working on a single piece at the same time. Processing time takes TRIA(10,12,15) minutes.
Second Process: There are two workers working on separate parts. They can work on different parts simultaneously. The processing time for each is NORM(3.1) minutes.
Third Process: There is only one worker for each piece. Processing time takes UNIF (8,20) minutes.
Workers cannot do each other’s work. The workshop will recruit a maximum of 3 workers over the existing system. Each worker in the first operation costs $400, each worker in the second operation costs $100, and each worker in the third operation costs $120.
We need to run the system for 5 working days over 12 hours a day and achieve the optimum result.
After creating our simple simulation model, which we discussed in detail above, we can now learn to use the Optquest tool.
First of all, after launching the Arena program, we can easily acces the Optquest for Arena tool from the tools section (Picture 1).
The opening screen of the Optquest tool is as in Picture 2. As we can establish a new mathematical model, we can also perform operations by selecting the models we have created before. Since we will learn how to do an optimization from scratch, we continue by choosing “new optimization” (Picture 2).
When we start a new optimization process, we see a tree structure on the left side of the screen (Picture 3).
Let’s examine this tree structure together so that we do not get confused when we go into details.
Optimization with Optquest Tool
There are a number of variables we want to control in our mathematical model. Based on the simulation model example we created, we want to control the number of workers in the system. “A maximum of 3 workers can be recruited in total, how should the recruitment of workers that will provide the most income be done?” Questions like these indicate that we will need to perform resource control.
We don’t always have to do a resource control here. We can control every variable structure and give lower and upper bounds. In our example, resources are the only structures that can be controlled, so only resources are included in the tree structure.
The checklist in our example is as in Picture 4. We also added the amount of our resources working in 3 different processes to our checklist.
In addition to the controls we will provide, there will be features on which we will observe the changes. If we look at our example, we are looking for an answer to the question: “How does the change in the amount of resources affect the total profit?” The total gain here is exactly the result we need.
We may want to observe the response of many values, such as the time that entities spend in the system, queue lengths, etc. The definition, we will create in the example we have is shown in Picture 5.
The sentence in our example, “A maximum of 3 workers will be recruited,” will be the constraint of our mathematical model. We will define such constraints in this section.
The view of the constraint we created in our current example regarding the total resource capacity is as in Picture 6.
I have shared the detail view of this constraint we created, with you in Picture 7.
In this section we will define the objective function with which we aim to optimize the total gain. Like maximizing, minimization-oriented objective functions can also be set up.
The objective function we created and its details are shown in Figure 8 and Figure 9.
- Suggested Solutions
If there are suggested scenarios related to the controllable values we choose, we define them in this section. In our example, there is no suggested scenario, so we will not use this part.
For example, if we had a suggestion that there should be 3 employees in the first processing station, we would enter this information into the system in this section.
This is the part where optimization settings are adjusted. Since our current goal here is to learn the optquest tool in general terms, I will continue without mentioning the details in order not to confuse you.
Optimization Results: Optquest for Arena
We have provided all the necessary definitions for our mathematical model. Now we can start the optimization process. When we start the optimization process, our simulation model runs in the background for all the lower and upper limits of our controls and the optimum result provided by each scenario is seen (Figure 10).
Let’s examine the results of the Optquest tool together (Figure 11).
In the scenario where the optimum profit is achieved, we see that 4 people should work in the first process, 1 in the second process, and 3 people in the third process. In this scenario, the total earnings were calculated as $8675.
The closest revenue to the optimum result was determined as $7670. We see that there are 3 different scenarios that provide this revenue. These scenarios are as follows;
- 5 people working in the first process, 1 person in the second process and 2 people in the third process
- 4 people working in the first process, 2 people in the second process and 2 people in the third process
- 4 people working in the first process, 1 in the second process and 2 people in the third process
In these 3 scenarios, the business makes a total of $7670.
In the current system, 5 workers were already working and a maximum of 3 workers could be recruited. When we look at the results, we observe that 8 workers which is the upper limit was not exceeded. We can see that our constraints are applied correctly.
Using the Optquest tool, we can perform optimization operations on much more complex simulation models. In this article, I wanted to tell you how to use this tool and analyze the result with a very simple example. You can express your opinions and questions. See you in the next article, regards.