The HV Battery Fitness sheet gives an overview of the battery block behavior
For every battery block, the following parameters are reported:
Avg Bias: every block value is compared to the max value of the whole battery. The difference of the block value from the max value is called Bias. This is the average bias over all the trip. This should be as little as possible. High bias may be a clue of a faulty block
Max Bias: maximum observed bias in the whole trip. A high value suggest the block is fluctuating more that the others
Highest Voltage: highest voltage the block has reached. At least one faulty block has been observed having both the highest and the lowest value of the whole battery
Lowest Voltage: lowest voltage the block has reached. Deteriotated blocks may have lower voltage than the others
Std Dev: standard deviation from average value: Blocks that fluctuate more have higher deviation. This is another sign of anomalous behavior.
Avg Local Delta: each block value is compared to its neighbours
Max Bias @stress: this bias is measured only when the battery is used with high current values. This should give a measure of how well the block behave under load.
Max Resistance: maximum measured resistance. Degraded cells may have higher resistance.
%Time @MinV: Time in the whole trip, the block was the one with the minimum voltage. This parameter filtered on delta above 0.2V, as lower values are considered not significant.
HSD3: "Batt Pack Current Val" "Battery Block Voltage -V01" "Battery Block Voltage -V02" "Battery Block Voltage -V03" "Battery Block Voltage -V04" "Battery Block Voltage -V05" "Battery Block Voltage -V06" "Battery Block Voltage -V07" "Battery Block Voltage -V08" "Battery Block Voltage -V09" "Battery Block Voltage -V10" "Battery Block Voltage -V11" "Battery Block Voltage -V12" "Battery Block Voltage -V13" "Battery Block Voltage -V14"
Android programmers wanted!
Torque Log Analyzer has been appreciated by several dedicated hybrid drivers around the world, and I'm willing to try the next step: bringing the level of analysis featured by my Log Analyzer to a real time, on-the-road application.
While I'm able to do some programming, I'm completely out of the Android development world, so I'm searching for an experienced developer willing to join forces on this project.
Here is a brief statement of work: if you're interested, get in touch with me.
Abstract The standard dashboard of Toyota Hybrid cars show only minimal information about the car status; the base version of the new Yaris even lacks the battery level indicator. An Android app can show car parameters that are missing from the standard cockpit with the aim of improving your driving.
High level draft The app should read data from the CAN bus of the car by using a standard Bluetooth adapter. The app may be a Torque plugin or standalone, if it has the ability to read CAN bus directly. In case of a Torque plugin, the app should be able to add a predefined set of custom PIDs to the ones Torque already has. PIDs and trigger values are specific to each car model, PIDs for Yaris, Auris and Prius are well known and already available. For the ease of user experience, there should be an app version for each supported car. Those versions may come from the same source code, just compiled with different defined constants.
Results The app will display at least two sets of values. Real time values:
Battery charge value
Engine coolant temperature
Current from battery
Brake usage (friction pads or regenerative)
Statistical (or cumulative) values:
% time of engine off
Km run in EV mode
Number of brakings
Code license I'm not a professional programmer and I'm not planning to earn significant money from this. As this is just a hobby for me, I favor the option of going open source for this project and releasing it for free on the market. Different options may be considered. If a paid version is considered, I'm willing to evaluate how to share revenues based on the contribution of each of the parties involved.
It's my opinion that, for the sake of simplicity, the current Toyota Hybrid cockpit is missing some valuable information for the driver.
Even the touch screen, that would be the perfect place to show more data, maybe out of an advanced mode, has little use.
Here is a Torque layout that shows what I consider the minimum useful info:
RPM: see if you're running on petrol or on battery, and when you're riding a sweet spot.
Battery State of Charge: as the Yaris charges the battery during the S1 phase, make sure you don't park the car with a full battery.
Battery Current: useful during the S1 phase, to check if you're demanding too much power from the battery.
Wheel Cylinder Pressure: check if you're braking using pads and thus wasting energy that would be recovered by the regenerative braking system. Note that the chart widget shows 10 seconds of history, so you can check braking performance later, without putting yourself in danger while driving.
Coolant Temperature: the S1 phase lasts at least until 40, and this guage will help you measure effects of winter radiator cover.
Google Sheet charts are handy but limited in scope: there is no easy way to produce a 3D chart.
3D charts can help discover relationships with multiple variables.
We'll use ParaView, an open-source, multi-platform data analysis and visualization application.
As an example we can use the Hybrid System Indicator:
The power meter is indeed related to the accelerator pedal but, as the experience suggests, its behavior seems to be affected by the SOC level also.
We'll use with a trip of 43Km with a SOC excursion from 38 to 63.
The analysis will be limited to the 0-100 HSI range, we'll exclude the regenerative braking range.
First, here are the values, as a function of the accelerator pedal:
Values seems to be limited by a minimum and a maximum straight lines.
Some values are outside this area: they may be transition values, when the accelerator was just pressed and the HSI indicator was still settling on the target value.
We can try to filter out those values by considering only HSI values associated to accelerator pedal values that are constant for at least 2 seconds.
The resulting chart is cleaner.
To add SOC values to the chart, we must switch to 3D with ParaView.
Export the Data sheet of Torque Log Analyzer in CSV format.
Note: you can't use the Torque log file directly, as the missing values will cause the data to be marked as text instead of numbers.
Torque Log Analyzer corrects those values when running an analysis, so just export a processed log file, not a raw one.
Open the CSV file in ParaView.
Add a Table To Points filter and select which column to use as X and Y (the same from the 2D chart) and Z (SOC, that wasn't used before)
X: Accel Pedal Pos
Adjust scaling and set Coloring at Batt Pack Current Val
Here's the chart:
Points color is related to battery current: charging current in blue, discharging in red.
Most of the blue points are in the upper part of the chart, where the accelerator pressure is higher.
There are really few red points (electric contribution to the car movement), but we must remember that these point represents constant accelerator values, not spikes where the electric engine is more likely to set in.
A little animation gives a better view of the data.
It's a common misconception that hybrid cars are not suited for highways.
Some people says the electric engine won't even kick in.
That's not correct, as the electric engine is nearly always contributing to the car movement.
Here's a chart of a 220 km highway trip: the red line shows how many amperes are flowing out of the HV battery, thus flowing through the electric engine and helping the petrol engine.
I had the chance to test a car with a little diesel engine (Opel Corsa 1.3 TDCI 95cv) on one of my daily urban commute trips.
The same trip was driven first with the Yaris Hybrid and then with the Corsa, at the same time of the day in 2 consecutive weekdays.
On this road, with the Yaris Hybrid, I have a personal record of 3.5 L/100km, attained in optimal conditions, in the evening, with no traffic, but the recording I'm going to analyze refers to a less than stellar 4.2 L/100km.
On a trip of 4 km, the engine was off for 63.56% of the time, with a total of 2.23 km driven with the electric engine alone.
On the same track, the diesel engine scored 5.2 L/100km, with a comfort and a noise far worse than the hybrid car.
The diesel engine was equipped with a start-stop system, but it never kicked in.
I suspect the A/C was preventing the engine from stopping.
This is another notable difference with the hybrid, that doesn't suffer from this limit.
There are some others different behaviors of the engine:
On the diesel car, RPMs of the engine are mostly proportional to the car speed.
On the hybrid car, this is not the case.
In the same chart, the engine load is also present: the hybrid engine always works with an high load.
The diesel engine, especially when decelerating, has a low load.
The average load value for the hybrid is 84.63% versus 53.35% of the diesel.
Before becoming an Add-on, Torque Log Analyzer was available as a stand-alone sheet.
This page contains instructions about using it, mainly for historical purposes: this is not the preferred way of using it.
Torque Log Analyzer is still available as a stand-alone script in the download section of this site.
Note that you can view the full source code also, so it may be useful if you're learning Google Script.
Anyone interested can go to the Downloads section and get Torque Log Analyzer.
Make a copy of the spreadsheet into your own Google Drive.
Open your fresh copy.
Select from menu "File / Import"
Select "Replace current sheet" and then select your trackLog.csv file Check the preview and then select "Import".
Select "OK" and "Accept" the following authorization requests
The sheet will be replaced by the content of your log file.
Select analytics and then click "Analyze".
If the sidebar doesn't appear, select "Start" from the "Add-ons" menu.
Spreadsheet will be populated with new columns and new sheet containing analysis and statistics about the trip you logged in trackLog.csv.
Also check for the "Log" sheet for any errors.
The Current Flow Chart shows electric current in and out of the traction battery.
With negative values, the battery is charging from braking, coasting or the petrol engine (see spot 1), while with positive values, current is actually flowing out of the battery for traction (see spot 2).
By using the EV Behaviour and the SOC Discharge charts, you can actually see where the current is coming from or going to.
If you switched to the new Google Drive, you may have experienced strange results.
It seems that new spreadsheets created in the new drive don't have a default locale setting.
When you load a CSV log file, first check "File / Spreadsheet settings..." and make sure to have "United States" selected: this is mandatory for decimal numbers usage.