Week 5 Battery characteristics using drive cycle

Week 5 Challenge - Battery Characteristics using Drive Cycle

Aim: To simulate a generic battery model's charge discharge using the UDDS drive cycle data

Theory

A drive cycle is a data set that shows how the speed of a vehicle varies with time elapsed. This data can be plotted on a graph with the speed on the y axis against the time on the x axis and this will give a more visual representation of how the speed varies with time. Drive cycles are used by automotive companies to test their vehicles on their performance such as mileage, emissions etc. Drive cycles are also important because they allow automotive companies to determine how much the power requirement required for driving HEVs and EVs based on various driving situations. By knowing the power requirement, they can select the appropriate energy storage systems that are capable of providing the necessary power to the vehicle. There are various drive cycles that follow the industry standard and each one is meant to simulate a particular driving condition that the average user will experience. For instance, the UDDS (Urban Dynamometer Driving Schedule) is a drive cycle that is a dynamometer test that simulates city driving situations for light duty vehicles.

UDDS Drive Cycle

For this assignment, we are required to simulate a generic battery model's discharge using the UDDS drive cycle data. The plot of the UDDS drive cycle is shown below:

SIMULINK Model

In this assignment, we will be using SIMULINK and the required blocks to model the discharge of a battery based on the drive cycle inputs. In the SIMULINK workspace, we need to insert the generic lithium ion battery block. The block is shown below:

If we double click the block, the block parameters dialog box opens and we can see the various settings for modelling the battery. This is shown below:

The image above shows the first main parameters of the battery block. As you can see, I have chosen the Lithium Ion type of battery, with a nominal voltage of 7.2V, a rated capacity 5.4Ah, an initial state of charge of 100%, and a battery response time of 30 seconds. I am not simulating the battery for the temperature and the aging affects. Similarly, I am not modelling any discharge parameters manually. Instead, I have selected the option for the discharge settings to be determined from the nominal parameters of the battery.

In order for us to model the battery's discharge with the UDDS data, we need to insert the drive cycle data in to the SIMULINK workspace. However, the data in the drive cycle shows the speed and the time, and the battery requires data in the form of currents. In other words, the battery block requires input data of currents that correspond to the driving pattern of the UDDS drive cycle. The spreadsheet of the UDDS drive cycle data for the speed vs time and current vs time have been attached to this assignment.

To insert the data set of the currents in to the SIMULINK workspace, I have used the signal builder block. This is shown below:

If we double click the signal builder block, the dialog box for creating a signal opens. This is shown below:

The signal shown in the above image is the default signal that is created when a signal builder block is added to the workspace. We need to add the current and time data from a separate spreadsheet in to the signal builder block. This is done by going to File > Import from File. Here, another dialog box will open and this is shown below:

In this dialog box, we need to Browse for the spreadsheet file which contains the current data for the respective speeds of the UDDS drive cycle. It is important that the data in the spreadsheet be organized such that the time data is in the first column and the current data is in the second column. It is also important that the spreadsheet be in the '.xlsx' format. Once the spreadsheet has been imported, select the tick box of 'Sheet1'. Finally, in the 'Placement for Selected Data' select the option, 'Append selected signals to all groups'. Click 'OK'. Now, we should be able to see that the current data has been added to the signal builder. By adjusting the time range to the length of the UDDS drive cycle i.e. 1367 seconds, and doing the same for the current values, we can see the overall current signal for the entire duration of the UDDS drive cycle. This is shown in the image below:

Next, we need to add the remaining blocks in to the SIMULINK workspace to complete the working of the battery simulation. First, we need a controlled current source block. This is shown below:

This block is used to convert an input SIMULINK signal in to an equivalent current signal. The generated current is driven by the input signal of the block. A connection should be made from the output of the signal builder block and to the input port 'S' of the controller current source block. The negative and positive terminals of the controlled current source will be connected to the positive and negative terminals of the generic battery block respectively.

The output port 'm' of the generic battery block can output multiple data from the input given. For instance, the battery block can give us the voltage, current and state of charge (SOC) data based on the input current values from the signal builder block. However, in order to control the observations from the 'm' port, we need to use a 'Bus Selector' which can process the data output from the 'm' port. For instance, if we want to observe multiple output data from one input, a 'Bus Selector' is used. The 'Bus Selector' block is shown below:

Once the bus selector has been added to the workspace, a connection should be created from the 'm' port of the battery block and the input port of the bus selector. We can customize the bus selector to output as many data as we want. For this assignment, I want to observe 3 main variables that are output from the battery block. These are the voltage, current and state of charge. To select these observations, we need to open the 'Bus Selector' block parameters. The dialog box that is opened is shown below:

From this dialog box, we can see that the bus has received the signals of the SOC, current and voltage from the battery block. So, I have selected each signal separately and given the 'Select >>' option to be added in the 'Selected signals'. Once all three signals have been added, we can click 'OK'. By selecting 3 signals, the 'Bus Selector' block will gain 3 output signal ports which we can use to connect with a 'Scope' block. The 'Scope' block will display the signals input in to it in a plot. The 'Scope' block is shown below:

Finally, we need to add the 'powergui' block in to the workspace since we are running a simulation with a continuous data type. The 'powergui' block is shown below:

The image below shows all the blocks and connections that need to be made in order for the simulation to run without any errors or warnings. We also need to modify the time for which we want to run the simulation for. In my case, I have run the simulation for the full duration of the UDDS drive cycle - 1369 seconds.

Opening the 'Scope' block will show the outputs of the SOC, current and voltage. These are shown below:

State of Charge (SOC)

From the SOC results, we can see that the battery started with an SOC of 100% since we had set the battery block parameters to have an initial SOC of 100%. This SOC level is maintained till t = 20 seconds because the vehicle is at rest for this duration. As the time elapsed, the SOC undergoes a general decrease from 100% at the start of the drive cycle to 95.77% at the end of the drive cycle. During the drive cycle, we can see that the SOC had also increased on several instances. For example, at t = 114.6 seconds, the SOC was 99.51% and at t = 123 seconds, the SOC has risen to 99.63%. This increase in the SOC is indicative of the regenerative braking phenomenon. These instances of the SOC increasing represent the vehicle speed decreasing i.e. braking. When the vehicle brakes, the mechanical work done by the turning of the wheels is used to generate electricity via the generator effect. The generated electricity is sent back to the battery for storage for later use and this is what we observe as the SOC increasing on the plot.

The steepest decrease in the SOC occurs at t = 187 seconds where the SOC was 99.49% and this fell to an SOC of 98.22% by t = 282.5 seconds. This corresponds to the steep increase in the velocity of the vehicle from the UDDS drive cycle. The high acceleration of the vehicle caused it to drain more electrical energy from the battery causing the SOC to decrease in such a steep manner. There were also several instances of when the SOC remained constant for a few seconds. For example, at t = 428 seconds, the SOC was 98.12% and t = 449 seconds, the SOC had not changed from 98.12%. This perfectly corresponds to a portion of the UDDS drive cycle plot, where the velocity of the vehicle was 0 m/s. This is seen also at t = 619 seconds to t = 647 seconds where the SOC remained the same at 97.6%.

There are several other instances of when the SOC increased or stayed the same, all of which corresponds to the respective motion of the vehicle. Overall, more energy was being drained from the battery than being recharged. In spite of this, we can say that the battery performed well for this particular drive cycle because the net decrease in the SOC was less than 6% for a total duration of 1369 seconds and was suitable for this particular driving condition.

Current

The current vs. time plot is exactly as expected in that it follows the signal that was imported in to the signal builder. We can see that in general, for the whole UDDS drive cycle, the current fluctuates continuously in both the positive and negative directions. Initially, for the first 20 seconds, the current flowing from the battery is 0A and this is due to the fact that the vehicle is at rest for this time period. This is evident during other instances, for example, from t = 125 seconds to t = 163 seconds, the current remained at 0A which again matches the drive cycle in that the vehicle was at rest during that period. When the vehicle needs to accelerate or maintain a constant speed, it will consume power from the battery causing its SOC to decrease over time. This is evident after t = 20 seconds, when the current shoots up rapidly from 0A to 4.28A at t = 31 seconds which corresponds to the vehicle accelerating from rest to a speed of 31 m/s. This shooting up of the current occurs many times throughout the entire drive cycle and it had also reversed its direction. For instance, at t = 114 seconds, the current had decreased to approximately 0A and continued to decrease in to the negative axis reaching an overall minimum current of-4.49A at t = 116 seconds. This reverse in the current direction is due to the regenerative braking phenomenon. Since electricity is generated by this process, it is then transported back to the battery for storage for later use and this flow is in the opposite direction to the current that is demanded by the motor. At t = 126 seconds, the current increases back to 0A and remains there till t = 163 seconds, showing that regenerative braking is over and that the vehicle is at rest.

The steepest increase of the current was when the current increased dramatically from 0A at t = 187 seconds to an overall maximum current of 8.1A at t = 195 seconds. This steep increase in the current corresponds to the increase in the velocity of the vehicle from 7.7 m/s to 16m/s. This was the greatest acceleration of the vehicle for the UDDS drive cycle.

Voltage

From the voltage vs. time plot, we can see that the voltage of the battery fluctuates throughout the UDDS drive cycle, however in general, there is a decrease in the voltage as the time elapses. Initially, the voltage begins at 8.412V at the start of the drive cycle and remains at that value changing very insignificantly till t = 20 seconds. This is the maximum voltage of the battery throughout the entire drive cycle. This constant voltage level is indicative that the battery is not being used for power, meaning that the vehicle is at rest. This matches the UDDS drive cycle where the vehicle was at rest during the first 20 seconds. When the vehicle's velocity began to increase, the voltage began to decrease. This is the opposite of what occurred in the current vs. time plot, where the current increased whenever the vehicle accelerated. From this, we can see that the voltage and the current are inversely proportional which is as expected. The voltage decreases to 8.337V at t = 26 seconds before increasing to 8.365V at t = 27.06 seconds. For the same time period, the velocity of the vehicle increased from 7.5 m/s to 7.7 m/s and the current in the battery decreased from 3.4A to 2.6A.

The second highest voltage in the battery achieved was 8.40V at t = 122 seconds. At this time, the velocity of the vehicle was decreasing i.e. it was braking and had a velocity of 3.93 m/s whereas, the current was negative being at -1.8A, yet it was increasing back to the positive axis (regenerative braking). Unlike the other two plots, the voltage vs. time plot has almost no observable instances where the voltage of the battery remained constant except for the first 20 seconds of the drive cycle. There is period where the voltage across the battery levels off after a sudden decrease from the 8.40V at t = 122 seconds to 8.288V at t = 163.3 seconds. It is during this period, that the voltage of the battery comes as close to being flat since the initial 20 seconds of the drive cycle.

Conclusion

In conclusion, the simulation of the generic battery model's discharge has been performed using the UDDS drive cycle data with SIMULINK. The results of the battery's voltage, current and state of charge have been plotted and the observations have been explained as well.