US20050116117A1 - Methods and apparatus for truck hunting determination - Google Patents

Abstract

Truck hunting determination involves taking samples of side to side acceleration in a vehicle, evaluating the samples by comparing the samples, and determining the degree of truck hunting on the basis of the comparison.

Description

RELATED APPLICATIONS
• This application is a continuation of application Ser. No. 09/119,461 filed Jul. 19, 1998, which claims priority of Provisional Application Ser. No. 60/053,143 filed Jul. 18, 1997. Applicant claims the priority benefit under 35 USC 120 of each of these applications.
FIELD OF THE INVENTION
• This invention relates to methods and apparatus for determination of truck hunting.
BACKGROUND OF THE INVENTION
• The various means for determination of truck hunting, that is lateral movement, in railroad cars and other cargo carriers have depended on cumbersome and inaccurate arrangements which are difficult to transmit to remote locations.
SUMMARY OF THE INVENTION
• One embodiment of the invention involves taking samples of side to side acceleration in a vehicle, evaluating the samples by comparing the samples, and determining the degree of truck hunting on the basis of the comparison.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1 is a schematic elevation of an embodiment of the invention in the form of a railroad car including equipment for truck hunting determination.
• FIG. 2 is a schematic elevation of an embodiment of the invention showing a vehicle roof line with a solar panel.
• FIGS. 3 and 4 are graphs illustrating performance under various conditions.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
• 1. Truck Hunting Determination
• 1.1. Definition
• Truck hunting is defined as a lateral movement from side to side of a railcar. The AAR has set parameters for defining problematic truck hunting as 10 or more events of greater than 0.75 g at a rate of 2-6 Hz. Lower levels of events are indicative of degradation of railcar components, and could further degrade.
• 1.2. Determination
• The aforementioned shaking leaves a measurable lateral signature on the y-axis accelerometer. This signature will be captured and monitored for evaluation under two different scenarios. The first deals with routine car health evaluation and monitoring, the second with truck hunting “events” in which the hunting exceeds certain levels.
• 1.2.1. Identification of Truck Hunting Events
• Two methods are available to capture this problem. They are:
• 1.2.1.1. Routine Monitoring
• During the course of operation, the unit will be taking samples and evaluating them (see below). The comparison of energy and peaks over time is one way to determine the level of truck hunting.
• 1.2.1.2. Event Monitoring
• The second method involves evaluating specific events that exceed certain set parameters, In this case, should the peak g exceed this number, then further analysis is done to determine whether it is a hunting event, or the car has gone over a piece of bad track.
• The unit will take the 3 second buffer and an additional continuous 7 seconds (for a total of ten) and analyze the number of peaks over a set g to determine if there has been a hunting event, or something else.
• 1.2.2. Truck Hunting Health Evaluation and Monitoring
• 10 second y-axis samples will be taken periodically (2-4 times per hour) throughout the day, and only when the car is moving.
• 1.2.2.1. Sample Analysis
• Each sample will have a number of peaks during the sample period. These peak numbers will be captured and stored in a database for eventual use in the truck hunting histogram. The data itself will be stored in the following format:
• Date/time, speed, mass, peaks, peak2, peak3, peak4, peak5, peak6
• Where:
• • Peak1=The number of peaks between 0.2-0.299 g
  • Peak2=The number of peaks between 0.3-0.399 g
  • Peak3=The number of peaks between 0.4-0.499 g
  • Peak4=The number of peaks between 0.5-0.599 g
  • Peak5=The number of peaks between 0.6-0.699 g
  • Peak6=The number of peaks at or above 0.7 g
• These bins will be stored on-car until the car is queried (once or twice a month) and then compiled into two histograms: one for unloaded, one for when the car is loaded.
• 1.2.2.2. Truck Hunting Histogram
• The histogram will have the number of peaks between these g levels within specific ranges of speed (see chart below).

  	0.2-0.299 g	0.3-0.399 g	0.4-0.499 g	0.5-0.599 g	0.6-0.699 g	0.7+ g

  10-19.9 mph
  20-29.9 mph
  30-39.9 mph
  40-49.9 mph
  50-59.9 mph
  60+ mph

Truck Hunting Histogram
• As the mass of the car may affect the hunting signature, at least two 36 bin histograms will be forwarded; one for when the car is unloaded, one for when the car is loaded. The addition of the two histograms together can all be done on the data server.
• 1.2.2.3. Trend Analysis
• The forwarded histograms will then be compared to the historical histograms we have on file. And significant changes can be noted and forwarded to the customer. The specific changes that might signal a trend towards truck hunting include any move towards the right columns (i.e. higher peak g's).
• 1.2.3. On-Car Parameters
• One of the benefits to the RM7000 system is the two way communication and the fact that much of the analysis and parameters are dynamic. We can, therefore, remotely set the key parameters as we learn more about the functioning of the cars. The key remote-settable parameters are:
• • Sample frequency—The number of times per day (e.g. 2 times/hour) that a sample is taken
  • Sample duration—The length of time (e.g. 10 seconds) that a sample lasts
  • Alarm trigger—The peak g that must be exceeded (e.g. 0.5 g) for capture to start and for analysis to be done on the captured sample.
  • Bin parameters—Histogram parameters (speed and g ranges)
    1.2.4. Commands
  • Reset—Will reset/re-calibrate the data on-car. This is needed if a repair has recently been done, etc.
    1.2.5. Messages
    1.2.6. Alarm—Wake
• The unit will automatically wake and capture data if y-axis peak g's exceed a certain level (i.e. 0.5 g). At this point the unit will save the buffer and capture for an additional period of time.
• This analysis will evaluate whether there have been a number of peaks over certain levels within the sample period of time. These parameters will be set in a manner similar to the AAR definitions. For example:
• If the number of peaks over 0.5 g is greater than some set number (5-10) then an alarm message is sent to the customer, alerting them to the potential problem.
• 1.2.7. Alarm—Badswitch
• If the peak exceeds some other number (e.g. 1 g) and seems to be a single event, then another alarm is sent, indicating that there was a bad switch, or perhaps a bad section of track or other anomaly.

  Summary of Truck Hunting Health Determination Event Flow

  During Standard Monitoring	After Awakening from Trigger
  Sample waveform (typically 10 seconds) is	Sample waveform is taken (3 seconds from
  taken	buffer, 7 additional seconds, for total of 10)
  Peak Analysis is performed (i.e. # peaks within	Peak Analysis is performed (i.e. # peaks wi
  each range of g [0.2-0.3, etc.])	each range of g [0.2-0.3, etc.])
  Results are saved in database
  Results are saved in histogram
  Waveform analyzed to determine whether the	Waveform analyzed to determine whether t
  frequency of significant peaks (i.e. those over	frequency of significant peaks (i.e. those ov
  some set number [typically 0.5 g]) is greater than	some set number [typically 0.5 g]) is greater
  some other set number (typically 2-6 Hz)	some other set number (typically 2-6 Hz)
  Histogram analyzed for trends to determine
  potential degradation of performance
  If trend alarm triggered, then alarm and
  histogram sent to base station
  If not, histogram forwarded at set period (e.g.
  month end)
  If any peak exceeds some set number (typically	If any peak exceeds some set number (typic
  0.5 g), and frequency is not greater than trigger,	0.5 g), and frequency is not greater than trig
  then alarm is sent to base station regarding	then alarm is sent to base station regarding
  possible bad track section	possible bad track section
  Periodic geographic analysis done in central	Periodic geographic analysis done in central
  server to determine any possible track	server to determine any possible track
  anomalies.	anomalies.

Claims (2)

1. A method of for truck hunting determination, comprising:

taking samples of side to side acceleration in a vehicle;

evaluating the samples by comparing the samples; and

determining the degree of truck hunting on the basis of the comparison.

2. An apparatus for truck hunting determination, comprising:

a side-to-side accelerometer;

a comparator responsive said accelerometer; and

an evaluator responsive to said comparator.

Images