Absimer-Recenl years progress in the Geld o f mlero-elcclromcehanieal nyrtcms (MEMS) itas nmde low cost inertial devices become n pouiblc opllon for land navigation. However, the wenk stand-alonc nccurncy and poor run-lo-run stablllty of low cost incrlial dcviccs innkc thcm not usable as sole navigation syslem. Even the intrgrntion of them as a supporting device into n navlgntion sysleni requires non-lraditional approaches and algorithms Tile objective o f this paper is lo assess Ihe fearlblllty of asing lhls kind of low cost inertial device, specificnlly the Crossbow DMU-6X. intcgrnted wlth GPS informalion, to provide lhe navigation cnpnbility to brldge the GPS outngei for tens of rrcondr. A series of lcsts were conducted in a land vehiele. Postprocessing rcsulls show lhal when GPS signal is available, the position nccurncy of low cost SlNSlGPS inlegration is jus1 ssme as lhnt of CYS only. With complete loss o f GPS signals, posllion error is shown lo he leis than 10 nielers, after 10 seconds. Prom these rcsulls, we can conclude lhnl ailhough CPS and current low cost IMU inlegration cannol improve absolute positioning areurncy, il has high data update rate and can give motion infonnnlion by SINS anly when GPS is no1 available for a short time, such ns in hcivily forested wens, in urban centers, In buildings or underground. b d c x Terius-GPS, IMU, Kalnisn Filter. Intelligent Trsnsport Syslcin I. INTKODUCTION A . arc widcly uscd in inililaty and civilian applications, such Although high accuracy incrtial navigation systcms (INS)as tlic navigation and altitude control of aircat? and spacecraft, its high cost limit its application in some general arcas, sucli as Ihc navigation of land vehiclc in intelligent transporrdlion systcm (ITS).Ncwcr low cost incrtial devices with comoact. lowerIn order to demonstrate the accuracy of the system. a series of tests were conducted in a land vehicle. Test runs were carried out in Singapore in Nov. 2001. The Crossbow DMU-6X and Sirf GPS receiver were installed in a car and driven throughout Pan Island Express, from NTU to Yishun reservoir.Post-proccssing results show that when GPS signal is available, the position accuracy of low cost SINSlGPS integration is just same as that of GPS only. With complete loss of GPS signals, position error is shown to be less than 10 meters, nfier 10 seconds. MATHEMATICMODELOFLOW COST SNS/GPs INTEORATION A. System slate equationThe application of the Kalman filtering technique to combine the inertial and GPS measurements requires adequate model for tbc SINS and GPS systems. The following error statc equation is applied in this paper:x ( t ) = F ( t ) X ( f ) + G(f)W(t)(1) where x(t) is 15x1 system stale vector, w(f) is 12x1 system noisc vector F ( f ) is 15x15 system dynamic matrix and C(t) is 15x12 system noisc matrix. X ( t ) is defined as follows (2)

A fiber optic gyroscope (FOG)-based measuring-while-drilling (MWD) device for horizontal drilling is developed and an in-field fast calibration method of the MWD inertial measurement unit (IMU) is presented. The IMU consists of a FOG triad and a quartz accelerometer triad. An error model for the inertial sensors is established and 12 errors are confirmed as the main error sources for horizontal drilling after an analysis. A five-rotation-in-level-plane (5RILP) in-field fast calibration scheme for the main errors of the MWD IMU is illustrated. The scheme can be conveniently implemented on an almost level plane, such as a table plane, without any special external equipment. To solve the calibrated errors, a systematic approach based on specific force measurement observation is adopted. Observation equations of the IMU errors are derived according to the rotation sequence of the calibration scheme. Simulation results show that the proposed calibration method is effective even if there are 5° errors in level and 10° errors in heading. Calibration experiments using the developed FOG-based MWD equipment on a table plane confirm the validity of the proposed in-field calibration method. Inertial navigation tests show that the actual accuracy of the MWD IMU is greatly improved by 60.6% after in-field calibration.

No abstract