Software Design and Development for Automotive ...

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

Software Design and Development for Automotive Online Monitoring System N.H. Norma#1, Z. Saad#2, I.S Isa#3, S. Ishak#4 #

Faculty of Electrical Engineering, Department of Science Computer and Mathematic1, Universiti Teknologi Mara (UiTM) MALAYSIA, 13500 Pmtg. Pauh, Pulau Pinang, MALAYSIA [email protected]

general purpose types to be programmable however many remain as static machines with only a limited number of variable parameters. Electronic Online Monitoring has replaced chart recorders in many applications. The vehicle online monitoring records the status data related to the security performance of automobile and integrate them, which can implement a real-time monitoring and diagnosis for the vehicle safety state. A lot of researches have been conducted on software implementation in data acquisition system. Research by Perez, Garcia, Nieto, Pedraza, Rodriguez, Zamorano (2010) [1] introduces the Advance In-Vehicle Data recorder (IVDR) to determine the most common reasons for Car accidents by studying the car driver behavior. This project is able to record the vehicle speed (vehicle data), the point of gaze (driver data) or the current distance to lateral road marks (environmental data). Argo is complex system built around a multicomputer platform based on image compression, video on disk recording, image processing algorithm as well as pattern-recognition algorithm. The system is fulfilled the established requirements and the expectation of the Spanish Traffic Agency. E.Larimore et al (2009) [2] present the identification and monitoring of automotive engines. The main objective is to extend and refine the nonlinear canonical variety analysis (NLCVA). The additional refinements are developed using general bases of non linear functions. The method of Leaps and Bounds with Akaike information criterion AIC are chosen. Delay estimation procedures are employed to consider the state order of the identified engine model and also reducing the number of estimated parameter that affects the identified model accuracy. The linear Gaussian system methods was applied to a 5.3L V8 engine to verify appropriate nonlinear basis functions, engine delay and decrease the model state from 16 to 7. Herpel et al (2009) [3] exposed a straightforward methodology how to perform prototype measurements on automotive CAN ECU communication and how to derive valuable information about controller-specific startup behavior. Logging and accessing important aspects of data transmission is done in both the early phase of system design by mean of simulation or analytical evaluation of the in-car communication system, and in terms of

Abstract –This paper presents the design and development of an online monitoring system that collects data over time or distance with a built-in instrument or sensors. Currently, an online data collection system (Vehicle Data logger) from general vehicle is not available in the market. Only the vehicles’ manufacturers have the tool to access the engine ECU to monitor the vehicle’s data. Current available automotives meter are display an estimated but inaccurate speed, engine revolution, fuel and temperature data. This project is developed to record the speed of vehicle, engine revolution (rpm), engine temperature, fuel volume and distance parameters. The software design of V-model has been developed to implement a sophisticated data viewing methods. The program utilized new framework based on track segmenting to better organize data, instantly provides summaries of segment data and attempts to better display the driving performance. The characteristic of fuel sender (to determine fuel volume in Litre) and temperature sender (to determine engine temperature in Celsius) is downloaded to EEPROM. All the recorded information are saved in RAM of the microcontroller, which can be reset after load the information to the (personal computer) PC using UART communication with Keyword 2000 protocol. All the measurements were carried out on selected road track as the field test of total trips about 3 kilometers. The recorded data have been analyzed and validated using an external (Global Positioning System) GPS navigator. Keywords – V-models, Data Logger, Bench Test, Vehicle Test

I.

INTRODUCTION

Most modern cars employ data acquisition systems that record many aspects of vehicle behavior. The amount of data recorded can be overwhelming, which can restrain the effectiveness of the data [9]. The purpose of this research is to increase the usefulness of the data through organizing, summarizing and analyzing. An online monitoring system is an electronic system that collects data over time or in relation to location either with a built in instrument or sensor or via external instruments and sensors. Increasingly, but not entirely, they are based on a digital processor (or computer). Online Monitoring vary between general purpose types for a range of measurement applications to very specific devices for measuring in one environment only. It is common for

978-1-61284-413-8/11/$26.00 ©2011 IEEE

489

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

is the verification phase and the right side is the validation phase. The verification phases contain a requirement analysis, system design, architecture design and module design.

measurements in real test cars by logging and analyzing communication data in prototype installation. The results from measurement data analysis show the comparison of CAN communication startup durations between the available ECU’s in the prototype CAR (German Car AUDI A6 Limousine 3.2 TDI). The advantages of those methods are the high speed CAN with real time data logging. Research by Changqing (2008) [4] explained about the LPT data acquisition card design of the front-wheel vehicle positioning test equipment based on the I2C Bus. With advanced I2C bus (smart internal bus) technology, the collected data in large capacity memory will send to PC via serial communication then PC will send to multiple devices data acquisition and processing system. The system has already used since past two years and the collection of data is still correct. Research by Brammer, (1997) [5] proposed to split the data logger into two main areas, vehicle performance monitoring and driver performance monitoring. The system is designed to be used easily and requires no input from the driver of the vehicle. The data logger records a set of data for each journey made in the vehicle where each journey starts when the ignition is switched on and ends when the ignition is switched off. The aims of this data logging unit is to connect to existing signals that are readily available in a modern vehicle. William Swihart and Jerry Woll, (1997) [6] have developed an integrated collision and vehicle information system for heavy machine. They have proposed a possible system integration path that would combine existing collision warning system (CWS) onboard computers (OBC) capability with next generation vehicle radar technology and emerging drowsy driver systems. The challenge to the systems integrator is to combine relevant data that provides utility to both the fleet operator and the driver without adding unnecessary system complexity and cost. II.

Fig. 1 The V-model of the System Engineering Process

B. Software Testing Technique The software testing technique is divided into two parts that are bench test for module and system test and vehicle test for real application test. The module test and system test are performed so that the error can be detected earlier and eliminated once the software development has finished. The module test will be performed at the module level development and system test will be performed at the system level development. The equipments used during the module and system test are oscilloscope, DC power supply, function generator and decade box. The test cases with the expected result are drafted before module and system test are performed. Before the vehicle test is conducted, all of sensor inputs on the wire hardness are verified to ensure the sensor input is correct connected to the online monitoring box. Oscilloscope and multimeter are used for the sensor measurement. The data collection from the input sensors of the online monitoring box is done in real time. All the collecting data were stored in PC memory.

METHODOLOGY

The objective of this study is to measure and analyze parameters from the developed online monitoring system. The automotive online monitoring system is designed with microcontroller based for data logging purposes and requires no input from the driver of the vehicle. The automotive online monitoring system records a set of data for each journey made in the vehicle which is starts when the ignition is switched on and ends when the ignition is switched off.

C. Hardware And Interfacing The earlier vehicles were implemented a conventional network topology. However disadvantage of this system is the undetectable sensor failure during the logging system. The data in the CAN BUS are still available in the network although the sensor is failed to provide an input. The integration of CAN network topology with conventional network topology are also valuable for middle class of vehicle. Most of current Malaysian Proton Car such as new Exora, Gen2, Persona and Satria Neo use this combination network topology.

A. Software Design and Development The V-model is applied for the software development process [8],[10]. The process steps are bent upwards after coding process to form the type V shape. The V-model demonstrates the relationships between each phase of the development life cycle and its linked phase of testing. The horizontal and vertical axes represents time or project completeness (left-to-right) and level of abstraction (coarsest-grain abstraction uppermost), respectively. The model is illustrated in Figure 1. On the left side of V-model

490

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

III. V

SUPPLY

POWER REGULATOR

DATA

This project encompass into two major important part of automotive online monitoring system that are hardware and software development. The hardware development describes about the microcontroller circuit and circuit interface from vehicle to the automotive online monitoring system. The software development explains the V-models use in the software design till the validation process. The online monitoring system is functioning to record the speed of vehicle, engine revolution (rpm), engine temperature, fuel volume and distance. The characteristic of Fuel sender (to determine fuel volume in Litre) and temperature sender (to determine engine temperature in Celsius) is downloaded to EEPROM. All the recorded information is saved in RAM of Microcontroller, which can be reset after load the information to the PC using UART communication with Keyword 2000 protocol. The summary of hardware and software development part is described in Table 1 and Table 2 respectively. The modules that were being considered upon developing the software and hardware parts are listed in Table 3.

EEPROM

Clock IGNITION DIAGNOSE

FREQUENCY INPUT MICROCONTROLLER

TACHO IN M-Factor pulses/rev

UART INTERFA CE DATA

SPEED IN K-Factor pulses/km

PC

FUEL SENSOR TEMP. SENSOR

Fig. 2 System architecture of the data logger

The conventional network topology was well-matched for the purposes of development of data logger for lower class Malaysian Perodua Kancil 660 and 850 cc car (above year 2002 manufacture). The Perodua Kancil was selected as a target vehicle because the engine is already using the electronic ECU as the instrument cluster. All sensors are integrated with ADC so that outputs are in digital form. If Proton Perdana, Waja, Wira and Iswara were selected as a target vehicle more work has to be done in the signal conditioning stage. These cars still use the mechanical type sensors to log the speed, rpm, fuel and temperature. So these types of vehicle are not practical for a simple and efficient electronic data logger. Base on the hardware design procedure and data input, the component of system architecture of the data logger is representing in Figure 3. The 8-bit Single-Chip Microcontroller is used as the central control unit for the data flow coordination. The decision to use this microcontroller is based on its capability to drive on a stepping motor for external meter control or cross coil. As a result the data logger design will be simpler, least time consuming and cost effective. The design process is initiated by first constructing the overall system specifications outlined in [7]. A 12V voltage supply is used to regulate a dc supply for microcontroller. The clock generates 8.00 MHz frequency to the Microcontroller. Four input that comes from difference sensor namely as Speed, Tacho, Fuel and Temperature. All sensors used are available in the target vehicle. The speed and tacho sensor will supply a fussy frequency that represents the value of speed and revolution per minutes (r.p.m). The fuel and temperature sensor will forethought the resistance values that indicate the current value of fuel left in the fuel tank and the recent engine temperature.

TABLE 1 HARDWARE DEVELOPMENT SPECIFICATIONS Phase Microcontroller Design

Microcontroller Microcontroller Configuration

Phase Verification

Validation

ECU 2

ECU 3

ECU 5

Specification Input/Output Analysis ROM /RAM Selection Interfacing (Input/Output/Protection) Input:Speed, RPM, Fuel,Temperature, Ignition, Battery Output: Serial UART – Keyword Protocol 2000 Controller: 8 bit NEC microcontroller ROM: 64k byte RAM: 2024 byte Input interfacing:Resistance as supply Interface protection: Thyristor and Zener Output interfacing: EPROM

TABLE 2 SOFTWARE DEVELOPMENT

Data logger ECU 4

ECU 1

PROPOSED SYSTEM

ECU n

High CAN BUS Low

Fig. 3 Data logger in-car with CAN network topology

491

Activities Requirement Analysis System Design Architecture Design Module Design Module Test Integration Test System Test (Bench Test and Vehicle Test)

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

TABLE 3 MODULES FOR THE SYSTEM DEVELOPMENT

where m-factor is the engine factor and r.p.m is revolutions per minute. The gauge response specifications based on frequency input signal versus m- factor of 8000 engine revolution type which is 2 pulses per revolution. A resistive sender provides the fuel tank level and the microprocessor measures the analogue input for the resultant voltage. Any increment of fuel level were not displayed during ignition is on (normal condition), but only be displayed when the ignition is off to on condition. The engine coolant temperature input is provided by a resistive sender and the microprocessor measures the analogue input for the resultant voltage. The input and output characteristics in Table 4 were programmed in the EEPROM.

Modules System States Container System Kernel Port Analog Port Digital Velocity Revolution Tank Input Coolant temperature Monitoring Travel Distance (Odometer) Calculation

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

The data logger operation is controlled by the microcontroller software code resident in the microcontroller. Additionally, the algorithms used to calculate the speed, revolution, fuel and temperature are presented in the paper by Z. Saad et al [7]. Table 4 is tabulated results measured in laboratory which were being used as the benchmark for the system validation.

IV.

Two tests were carried out are bench test and vehicle tests. The bench test is developed using Microsoft Visual Basic 2005 where the online monitoring data and the computer is interfaced to UART communication with baud rate 19,200 baud. The vehicle test was conducted around UiTM Kampus Pulau Pinang road track for about 3 kilimetres distance. Three types of data that are speed, engine revolution and travel distance were successfully logged into the notebook through the online monitoring data. The data reader program inside the notebook converted the data from the machine language into Microsoft Excel file. The results are pictorial in Figure 4 till Figure 8.

TABLE 4 MEASURED PARAMETERS (BENCHMARK)

Input Freq (Hz) 0 14.16 28.31 42.47 56.62 70.78 84.93 99.09 113.28 127.44

Tachco Gauge (Revolution)

Speed (km/h) 0 20 40 60 80 100 120 140 160 180

Input Freq (Hz) 0 25 50 75 100 125 150 175 200 -

Fuel

Revolution (r.p.m)

Fuel (Lit)

0 1000 2000 3000 4000 5000 6000 7000 8000 -

5 8 12.5 25 37.5 45 -

Resis tant (Ώ) 304 284 252 188 124 77 -

Temperature of Engine Coolant Resis Temp tant (˚C) (Ώ) 45 222.3 50 181.1 112 23.6 117 20.6 -

Velocity Versus Time 70 60 50 Velocity (Km/h)

Speedometer Gauge

The speed input is received from an open collector hall sensor [7]. The input signal is calibrated by means of the pulses/km number [k-factor] programmed into the EEPROM. The k-factor is stored in the EEPROM with a resolution of 1 pulse/km. K-factor storage is limited to values between 1000 and 20,000. The speed is calculated as (1). f × 3600 Speed = k factor

RESULTS AND DISCUSSION

40 30 20 10 0 1

17 33

49 65 81

97 113 129 145 161 177 193 209 225 241 257 273 289 305 321 Time (Second)

Fig. 4 Velocity versus time (seconds)

(1)

Revolution Versus Time 4000

where k-factor is 2548 pulse/km and f is the input frequency. For the Tacho gauge, microcontroller measures the frequency of the pulses received on the input and drives the stepper motor to a position dependent on the frequency. There shall be no visible ‘step’ movement of the Tacho gauge. The tacho revolution is calculated as (2):

3500

Revolution (R.P.M)

3000 2500 2000 1500 1000 500

f × 60 Tacho = = r.p..m m factor

0

(2)

1

17 33 49 65 81 97 113 129 145 161 177 193 209 225 241 257 273 289 305 321 Time (Second)

Fig. 5 Revolution versus time (seconds)

492

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

Trip Versus Time 3.5 3

Trip (Km)

2.5 2 1.5 1 0.5 0 1

17 33 49

65 81 97 113 129 145 161 177 193 209 225 241 257 273 289 305 321 Time (Sec)

Fig. 6 Trip (KM) versus time (seconds)

Fuel Voiume(Litre) Versus Time (Sec) 25

Fuel Volume (Litre)

24.99 24.98 24.97 24.96 24.95 24.94 0

50

100

150

200

250

300

350

Time (Sec)

Fig. 7 Fuel Volume (litres) versus time (seconds)

Fuel Voiume(Litre) Versus Trip(KM) 25

Fuel Volume (Litre)

24.99 24.98

pedal) is about 900 RPM to 1100 RPM with air condition on. The minimum speed was recorded at the second of 91, the speed was 2.8 KM/H at 1225 RPM and by the time the trip was about 0.6 KM. The maximum speed was achieved at the second of 237 the speed was 57.9 KM/H at the 3034 RPM. By the time the trip was about 2.1 KM. The fuel volume was fluctuated due to the tank sender float was going up and down due to the road is bumpy. The car consumed around 0.02 Litre for 3 km. The Engine Coolant temperature was remains stable at ADC value 153. The Coolant Fan is activated when the engine temperature start to increase. This analysis is done manually and technically from the driving experience and technical knowledge of vehicle engine. This data also has been validated with an External GPS Navigator. The differences between online monitoring data system to other data logging or instrument cluster is that the online monitoring data measured the fuel volume with high resolution 0.001 litre compare to the instrument cluster. The instrument cluster shows only Empty, Half and Full without indicates the actual fuel volume in Tank. The engine tuning can be performed using the online monitoring data by monitoring the idling condition. The real engine revolution input was sampling for every 1 second, therefore the engine condition can be monitored. The engine idling for Perodua kancil is around 1153 r.p.m when the Coolant fan is switching on. The fuel volume was measured from the sensor and convert to the Litre. So the fuel consumption can be monitored by travelling distance. The fuel was spending for 3 km is 0.02 Litre on the Uitm track. The tank sender float was going up and down on the bumpy track was shown during vehicle test. Therefore using data monitoring system can save a lot of time, energy resources, and money. The UART communication was successful applied for the data logging. During the data logging the online monitoring data remain communicate with the computer without any break. Data monitoring system used and existing sensors and system in vehicle without any modification while some conventional or other methods need a modification in order to measured the fuel consumption signal.

24.97

V. CONCLUSION

24.96

The V-model technique is very helpful for the software development. The bug can be minimized or eliminated by performing the module and bench test. It is also can expedite the time of software development. The NEC microcontroller can support for other application likes CAN Bus, LIN Bus and Serial communication likes Manchester code. The microcontroller has reserved more memory space for expend the application. The UART communication is still can be applied for the data logging even the automotive manufacturer is currently using CAN bus for the data communication likes Proton car. The Online monitoring data can save a lot of money and time to measure the data from the sensors. The behaviour of the sensor input can be monitored by using the online monitoring data. The existence of the online monitoring system has an advantage for the vehicle consumers. Since the system could retrieve

24.95 24.94 -0.5

0

0.5

1

1.5

2

2.5

3

3.5

Trip (KM)

Fig. 8 Fuel Volume (litres) versus trip (KM)

From the data speeds versus time as in Figure 4, it has exposed that the target vehicle was moving after 10 seconds the engine was started. But the oil pedal was pushed before that. The data RPM versus time as in Figure 5 can show clearer evidence at which second the driver start to push the oil pedal. Start from second 4 until second 9 it shows how drastically the RPM was increase from 1153 RPM to 2058 RPM. Normal RPM for engine without ramp (push the oil

493

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

accurate signals such as the fuel volume with resolution 0.01 Litre and the travelling distance in meter with resolution 100m, actual speed in km/h with resolution 0.1 km/h and actual revolution. The online monitoring data will be activated once the car switches is on and starts collect the data when the computer requests the data. Using the online monitoring data the driver can also detect the defective sensor when the sensor input remain at certain value or zero during car moving. The driver could realize the fuel consumption in seconds and can optimized their driving manoeuvre. The retrieved data from the sensor can be studied and analyzed for further improvement of the vehicle. The project is proposed to upgrade the online monitoring data instead of using UART communication and implement CAN communication as because the communication is faster than UART communication. The UART speed is 19.2 KBaud but the CAN communication can achieved until 1 MBaud. The CAN tool for CAN communication validate need to identified for the CAN development. Normally the CAR manufacturer or automotive supplier will use the CAN Vector tool for the CAN development but more budget need to be allocated due to expensive tool. REFERENCES [1]

Perez, A., Garcia, M.I., Nieto, M., Pedraza, J.L., Rodriguez, S., Zamorano, J, “ Argos: An Advanced In-Vehicle Data Recorder on a Massively Sensorized Vehicle for Car Driver Behavior Experimentation”, IEEE Transactions on Intelligent Transportation Systems, pp 463-473, 2010. [2] Larimore, W.E., “Maximum likelihood subspace identification for linear, nonlinear, and closed-loop systems”, American Control Conference, pp 2305-2319, 2005. [3] Herpel, T., Kloiber, B., German, R., Fey, S, “Assessing the CAN communication Startup Behavior of Automotive CAN ECUs by Prototype Measurement”, In Int Instrumentation and Measurement Technology Conference ( I2MTC) Singapore, 2005. [4] Changqing Cai, “The LPT Data Acquisition Card Design of the FrontWheel Vehicle Positioning Test Equipment Based on the I2C Bus”, International Symposium on Intelligent Information Technology Application Workshops, pp 753-756, 2008. [5] Brammer, A.J., “ Simple data logging system”, IEE Colloquium on Monitoring of Driver and Vehicle Performance (Digest No: 1997/122), 1997. [6] Swihart, W.R., Woll, J.D., “Integrated collision warning and vehicle information systems for heavy vehicles”, IEE Colloquium on Monitoring of Driver and Vehicle Performance, pp. 2/1 - 2/7, 1997. [7] Zuraidi Saad, Noor Hafizi Norma, Azhar Efendi Md Saad, Fairul Nazmie Osman, Khairul Azman Ahmad, “Design And Development of Universal Data Logger For Testing Vehicle Performance”, Conference On Scientific And Social Research (CSSR), Digest no: 570555 , 2009. [8] Christian, B, “ The V-Model Development Standard for IT-System of the Federal Republic of Germany”, IABG. Physical Review: http://www.v-modell.iabg.de, 1999. [9] Babu, S., Shashidhar, Y.M., “An Intelligent Online Mileage Indicator for new generation Automobiles”, IEEE International Conference on Industrial Technology, 2006. ICIT 2006, pp 2250 – 2255, 2006. [10] Johansson, C., Lecture notes of V-Model, Quality Management, DPT404, IDE, University of Karlskrona/Ronneby Wikipedia (2010). Physical review on V-Model (Software Development) from http://en.wikipedia.org/wiki/v-model_software_development., 1999.

494