CAN-View Data Analysis, Complete with Charts

CAN-View Data Analysis, Complete with Charts & Annotations

After having kept my malfunctioning CAN-View data capture on the back burner for a year or so, thanks to Jay Groh (aka TheForce) I finally got it working. For background on the process, see this. For more on what Jay is doing, see this.

I enjoy learning about this fascinating car. I am a technologically challenged visual learner, so charts are a great help to me. And to others, I'm sure. So I plan to give MS Excel's chart wizard a workout with my CV data, with results posted here and, where relevant, elsewhere.

This thread will be a work in progress as I grab more data and add new and different charts. My first goal is to help with my own understanding of how the car behaves under the hood during everyday driving and in certain fuel economy tests that I plan to perform, sharing my observations along the way.

In addition, if there are certain driving conditions and data you'd like to see presented, let me know. Charting additional data elements from previously presented drives should be no problem. For types of driving I haven't already logged, I'll be glad to consider them to the extent time allows and they aren't contrary to good fuel economy. For example, don't ask me to chart instantaneous MPG at 4000 RPM while climbing a steep hill at 75 MPH. If I happen to make a trip where such driving is required, I'll have the laptop running, but that's about it.

This is a partial list of what the data capture includes:

• Hybrid battery voltage
• Current flow into and out of the hybrid battery
• Temperatures of the coolest and warmest hybrid battery modules
• Speed
• ICE RPM
• ICE temp
• Coolant temp
• Accelerator pedal position
• Brake pedal position
• Shifter position
• EV mode status
• Cruise control status
• Headlight status
• Fuel injector timing (in some unknown units, as of now)
• Fuel flow
• VVTi (ie, intake valve timing in degrees relative to TDC)
• Fuel gauge (in percent full)

From there I have set up an Excel template that calculates these additional parameters:

• Approximate cumulative miles per data point
• MG1 RPM
• MG2 RPM
• Instantaneous MPG

So with that, my first chart (next post) ....

 

This is my regular afternoon commute, with an elevation profile:





The route can be seen here. The afternoon commute is south to north. Substantial pulse and glide is used.

I'll add some annotation, probably tomorrow -- I'm going to bed shortly. But I do want to explain one thing: The X axis shows mileage, but the data points are based on time increments, approximately one every two seconds. So mileage increments are not linear from start to finish; the faster I'm driving, the greater the increment.

I may need to play a little with chart formatting and resolution. I can see that it's hard to differentiate the markers for EV and brake in the legend.

EDIT/UPDATE on weather: 79F, sunny, light winds.

 

A chart from the same commute with current flow and SOC, along with EV and braking events:

 

Here’s the annotation for the commute by “mile markers.†Because of the previously mentioned discrepancy between time increments and the X-axis mileage increments, these markers may not match the elevation profile exactly. They are based on the spreadsheet from which the charts are generated, using this formula:

((Total trip mileage / total count of data points) * (speed / trip average speed)) + mileage increment for the previous data point.
​

The description is fairly detailed for the benefit of those who might compare it to the charts and who want to know details of driving strategy and technique.

I should mention that though I have an engine block heater, there is no preheating; the area around the office provides nowhere to plug in.

• 0.0-0.4: Through the office complex, good for low-speed, low-demand S1 operation. S1 generally ends at about mile 0.25, enabling EV mode activation. Segment ends with a stop sign and left turn.

• 0.4-1.2: Lightly traveled 25 MPH residential area, favorable for low-speed P&G. Ends with a stop sign and left turn.

• 1.2-3.9: Moderately traveled secondary road, 35 MPH, two lanes. Traffic volume prevents sustained P&G from start to finish, but short P&G segments usually can be worked in, often facilitated by gentle hills. Several short right-turn lanes along the way often allow a brief pull-off to allow faster traffic to pass. It is generally on this stretch that ICE temp reaches the 70C S3 threshold in warmer weather. Until then, however, traffic often forces speeds >34 MPH and prevents EV mode activation. To minimize frustrating ICE-on downhill glides, I try to come off a hill crest at <34 MPH while activating EV and immediately shifting to neutral to keep the ICE off downhill. The segment ends with a gentle uphill approaching a stop sign, allowing a nice terrain-mediated glide and slowdown. Right turn onto the next segment.

• 3.9-4.2: Another 35 MPH secondary road, more heavily traveled than the previous segment. Steady downhill, great for gliding but requiring a left turn at the bottom, with significant braking, sometimes to a complete stop to await a break in oncoming traffic. Oh, the waste of kinetic energy!

• 4.2-5.2: Moderately traveled 25 MPH residential cut-through. Begins with a steep uphill then levels out, allowing good P&G segments. Ends at a stoplight with a right turn, normally red but occasionally a green arrow. Occasionally I can sit long enough for the S4 idle check, but not usually.

• 5.2-6.2: 45 MPH thoroughfare, four lanes, moderate to heavy traffic. There is a light about ¼ mile into the segment which sometimes helps – if it’s red when I make the turn, I can justify keeping my speed down. Other times it hurts – a stale green that I know will be turning any time, but the tailgater behind me wanting to do 45+ right up to the last second doesn’t consider that possibility. On this day, it turned red shortly after I turned onto the main road. I had to apply light braking but never came to a complete stop before it went green. From there I generally can keep it about 35-40 MPH without disrupting flow. The segment ends with a left turn at another light. It’s often a challenge moving from the right lane across the left travel lane to the turn lane; there’s not always a break in traffic in the left lane that allows me to make the move without punching it up to 45. So plan B is to stay in the right lane, turn right at the light onto a short and very lightly traveled residential street, and turn around at the first driveway. It’s a short downhill back to the light that requires minimal EV. It is at this intersection that I usually trigger the S4 idle check if not done previously.

• 6.2-9.3: Moderately to heavily traveled secondary road, 35 MPH, two lanes. It begins with a short but steep drop and climb, then levels out for about a mile. Short bursts of narrow-range P&G (33-40 MPH) are often possible. But here is where one of the annoying idiosyncrancies of the Prius propulsion system often rears its ugly head: I can’t get the ICE to shut off in S4! Traffic usually is too heavy to reduce speeds for EV-button glides, so I’m forced to putter along in ICE-on, low-RPM conditions when external conditions would allow an ICE-off glide. At about mile 7.5 the road makes a 90-degree bend to the right. There begins a roller coaster ride of almost two miles with a net elevation increase. First is a gradual, fairly long downhill that grows progressively steeper. With no traffic behind after the bend I can glide and gain speed. I shift into neutral before 40 MPH to allow warp-neutral. Speeds can top 45 MPH at the bottom, so I often shift back into D to avoid a speeding ticket. The uphill approaches a stoplight, usually red. After the light, the hills are shorter but more numerous. Another light, also usually red, ends the segment. I keep straight at the light.

• 9.3-10.8: Now four lanes, still moderately traveled. The first 1/3 mile or so approaches the busiest intersection of the route, with a light that’s usually red. With the extra lane I generally can keep it below 30 MPH as I approach the light without impeding traffic. After the light, trip MPG drops considerably: Speed limit increases to 45 MPH, and it’s mostly a steady climb. Sometimes traffic is light enough to allow relatively slow P&G (25-35 MPH) on the (short) more level segments, but it’s mostly an ICE-on affair. A left turn at a stoplight ends it.

• 10.8-11.1: Two lanes, 35 MPH, a short uphill and a short downhill to the next (usually red) light, at which I turn right. I try to keep speed and RPM down on the climb, knowing I can glide after the crest and that I’ll have to stop shortly afterward. However, the motorists that often follow won’t look (physically or mentally) past my rear bumper, so I generally have to push it a little harder than I’d like.

• 11.1-12.2: Two lanes, 35 MPH, a short uphill then mostly downhill allowing a glide. I don’t want too much speed because I make a left turn into my development at the bottom. Traffic sometimes forces me to go little faster than I like, however. I can generally make the left turn with just a light brake check, though infrequently oncoming traffic requires a harder brake check. After the turn onto the development’s 25 MPH street, it’s a gentle but steady uphill gradually tapering off to near-level as I reach my house.

Charts, elevation profile, and annotation for the morning commute to come ....

 

Morning commute from 6/3. 12.2 miles & 78 MPG for the trip. OAT 57F with light winds.





To allow a direct comparison to the PM commute, the annotation uses the same mile marker identifiers for each segment, but, of course, in reverse order.

• 12.2-11.1: My street is nearly perfect for a cold-ICE, low-power, low-speed S1 warmup. As the current flow/SOC chart shows, I putter along generally with less than a 10-amp draw, with the road at first almost level and then gradually getting a bit steeper. Even in the dead of winter I’m in S2 before the right turn at the end. After the turn, there is usually no traffic behind, allowing a nice leisurely and ICE-efficient pace up the hill to the light, usually red. Left at the light.

• 11.1-10.8: Less traffic than in the p.m., so I often can crest the hill slowly enough to mostly glide to the right turn at the light, usually red.

• 10.8-9.3: A great mile-long glide! After making the turn from the previous segment I’m on a downhill. I pulse up to about 27-28 MPH, pop her into neutral, and enjoy the ride. That’s enough to carry me to the first light, usually red. Another short pulse after the light to carry me to the next one, also usually red. Only an occasional vehicle approaching behind. I’ll change lanes as needed to give them the one they want.

• 9.3-6.2: The two-lane roller coaster ride begins. I’m not warm enough for S3 yet so I want to keep my speed low enough for an EV button-shutdown before each downhill. The first part of the segment actually seems busier than the previous four-lane stretch. To reduce the chance of cars on my butt, as I go through the first intersection I’ll check behind for approaching vehicles, and glance at the cross street to the right for those that might turn in behind me. If there are either, there’s a short right turn lane into a drug store lot I can use to let them by before the first descent. Still, I usually end up with company on my rear before the next light. The RPM chart shows the highest engine speed of the trip through here as I tried to maintain close to 35 MPH on the uphill. Much of that traffic turns off in the middle of the segment or at the next light, so I’m usually unmolested for the long downhill and uphill before the 90-degree bend. On the level stretch following the bend, cars often pull in behind from adjacent subdivisions, though I often still can get some P&G segments. As an aside, the County’s Finest have been running radar through there quite a bit in the last few weeks. I’ve wondered to myself if or when they become suspicious of a 20 MPH-er passing through. The segment ends at the main road just past the short dip and climb.

• 6.2-5.2: The 45 MPH thoroughfare, nearly identical scenario to the afternoon trip other than somewhat lighter traffic: P&G when traffic allows; a light in the middle of the segment with a balancing act between timing it perfectly and courtesy to those behind me; a move across two lanes to turn left at the end; an optional right turn into a residential area if I can’t make the lane change; and an S4 idle check opportunity.

• 5.2-4.2: The residential cut-through, light traffic at this time of the morning allowing a great glide. It gets steep as it approaches a stop sign at the end, so I’m always on the brake before making the right turn. That’s where you see that burst of regen on the current/SOC chart. On a good day, no one’s behind me so I can approach the steep section at minimal speed – 15 MPH or so.

• 4.2-3.9: Now uphill on the more heavily traveled segment, sometimes requiring a hard push from near zero to 35, only to have to slow down again to turn left at the top.

• 3.9-1.2: I’m sharing this road with a few school buses this time of the morning. Most of them don’t impede me, and one of them actually helps from time to time. The first part of this segment is a moderate downhill to cross a creek, then uphill. About 2/3 of the way up the hill is a bus stop, with the bus coming the opposite way. If I’m lucky the bus is approaching or making its stop as I come down the hill and cross the creek – which it is with surprising frequency – giving me the excuse to continue gliding up the hill. If traffic follows and I miss the bus, so to speak, I’m pushing considerably harder up the hill. Good thing in either case is that the road starts leveling out at the bus stop. From there, I probably have about the same number of P&G opportunities as in the afternoon. There’s less traffic, but fewer right turn lanes to allow it to pass. Right turn onto the next segment.

• 01.2-0.4: Uneventful slow-speed P&G through the neighborhood.

• 0.4-0.0: A nice gentle glide with a short pulse here and there through the office complex and to the parking lot.

EDIT: Once again, there is no EBH preheating. Nowhere to plug in at my townhouse.

 

Thanks, Bob. I don't suppose it can be calculated from what is available, can it?

 

Anything relevant that is both in the data capture and decoded into CSV format is listed in my first post. I calculate MG1 RPM using (ICE RPM * 3.6) - (MG2 RPM * 2.6). MG2 RPM is calculated using MPH * 57.

This lists what is supposed to be available via the decoder. FWIW, contrary to what that page says, the decoder is not providing individual wheel speeds at all, and I'm getting values of zero for the accelerometer data. Not that I'm terribly interested in either of those for now.

EDIT: Brain cramp. I am getting wheel speeds.

 

OK, time to call on the brain trust that is PriusChat. Jay and I have had some dialog about these unknown units for fuel injector timing -- and we're not even sure that's what it is. They're labeled simply "Fuel Injector" in the decoded data, so that's our assumption. What else might it be?

Sticking with that assumption for the moment, we can't figure out what units they're in. If it truly is injector timing, it is many multiples of mS. That's fine if we can figure the multiplier. What's also puzzling is that they drop to zero very infrequently during ICE-off periods.

Bob, Hobbit, or anyone else who might know: Please take a look at this chart and give me your thoughts. These data are from a road trip this past weekend, with a mixture of all types of driving -- from low speed P&G through town to sustained highway driving, with a lot of mid-speed (40-50 MPH), low-RPM conditions on byways in between.

If this is injector timing, is its relationship with fuel flow consistent with what you would expect? If so, any guesses as to the multiplier? Finally, why does it return to zero so infrequently?

Are there other data elements to chart from among those available to help answer the questions?

Many thanks!

 

It will help me better understand the car and what's happening when I drive it. Hopefully it will help others understand it as well. And it appeals to my inner data geek.

Like you, I'm not sure how much I can do to improve fuel economy. But maybe the data will give me some hints. For example, I notice just with what I've already captured how much of a negative spike I get in current flow when I first hit the go-pedal. With the short duration, I'm sure the impact is small. But it's made me more aware of how I might improve that initial acceleration -- immediate substantial pedal pressure to light the ICE followed by promptly backing off to keep RPM from going through the roof.

 

A discussion on battery use in another thread raised the issue of how much current flows out of the battery during a no-arrows glide. This chart from one of my recent commutes shows the relationship between current flow and speed during ICE-off glides. Data are limited to glides with go-pedal positions of less than 10%.



Why does the current flow increase with speed?

I also looked at the relationship between pedal position and current flow. This chart includes events from the chart above and those with greater pedal pressure, whether from initial acceleration before the ICE lights, EV mode (non-switch activated), or an inadvertently heavy glide foot.



This seems to confirm what previous CAN-View observations of pedal position values have suggested, that a glide can be executed with pedal position anywhere from 1-10%. I did think there still was a relationship between pedal position and current flow during glides, but this seems to demonstrate otherwise until after the 10% threshold is crossed. And at that point, the car no longer is gliding anyway. Presumably it is above the 10% mark that the "no-arrows" condition disappears from the Energy monitor. But since my CV uses the factory display I can't monitor both simultaneously to verify.

 

I already have some data using N; I tend to use it during longer sustained glides. Like Jay says, it moves within a much narrower range, generally <1.5 amps. I'll see if I can get a chart put together in the next day or so. Pretty busy right at the moment, preparing for a flight to Las Vega tomorrow. :rockon:

Another planned test is a head-to-head comparison of gliding methods (N vs. pedal control) on a long gradual hill. I've been curious for some time about whether there is any added drag with N, as Jay suggests. The test will be simply to hit the top at a set speed, apply one glide method or the other, and document speed at the bottom.