CN102564689B - Zero correction method and device of torque sensor - Google Patents
Zero correction method and device of torque sensor Download PDFInfo
- Publication number
- CN102564689B CN102564689B CN201110449600.2A CN201110449600A CN102564689B CN 102564689 B CN102564689 B CN 102564689B CN 201110449600 A CN201110449600 A CN 201110449600A CN 102564689 B CN102564689 B CN 102564689B
- Authority
- CN
- China
- Prior art keywords
- value
- zero point
- reference value
- zero
- torque sensor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Power Steering Mechanism (AREA)
- Steering Control In Accordance With Driving Conditions (AREA)
Abstract
The invention discloses a zero correction method and a device of a torque sensor. The method comprises static zero learning and dynamic zero correction, and mutually checks and conforms according to a certain time sequence and an algorithm, so as to filter off disturbance and guarantee the control accuracy, and can dynamically track zero deviation in real time during the using process of a user. The method limits the control algorithm within a limiting range of the torque sensor, seriously guarantees the safety of zero correction, ensures that a steering system is always symmetrical and balanced during the driving process of the user, and even if zero correction range is exceeded to a certain degree, a controller can immediately cut off power assisting output and light up a fault light, provides driving comfort and safety for the user to the maximum degrees, ensures that the electric power assisting steering system is optimally matched, and prolongs the service life of products.
Description
Technical field
The present invention relates to a kind of bearing calibration and device of torque sensor, relate in particular to Zero correction method and the device of torque sensor in a kind of automobile electric booster steering system.
Background technology
In prior art, vehicle steering is with electric boosting steering system, work as steering wheel rotation, when moment of torsion is passed to torque arm, there is deviation in the sense of rotation between input shaft and output shaft, torque sensor is responded to this deviation, in conjunction with the signal of vehicle speed sensor output, determine the size and Orientation of power steering again, and utilize the operation of drive motor assisted diversion.
The long-term use procedure of electric boosting steering system meeting causes that torque sensor installation site is offset, torque sensor potentiometer spring restoring force is aging, torque sensor potentiometer slide block produces mechanical wear etc. with keeping a grip on spiral ball groove, so that the mechanical zero of electric power steering moment of torsion and deviation occurs electric zero point.Make driver occur that left and right turn moment is asymmetric when steering dish, the ability of vehicle loses straight-line travelling, can cause active steering or direction to be killed when serious.Prior art only just can be carried out zero correction when user keeps in repair, and while normally using for user, this function lacks completely.
Summary of the invention
The invention provides the Zero correction method of torque sensor in a kind of electric boosting steering system, utilize automatically controlled algorithm to make up the aging of physical construction, guarantee comfortableness and security that user drives, and extend as much as possible service time of product, make whole electric boosting steering system reach balance.
A Zero correction method for torque sensor, comprises the steps:
(1) static study at zero point:
A) the front reference value at zero point of confirming after working on power several times and preserving that reads respectively torque sensor obtains reference value queue at zero point.
Torque sensor after working on power each time the reference value at zero point during all by this end-of-job preserve, after repeatedly working, will form reference value queue at zero point, can guarantee that zero point, reference value was in use accurately and reliably transmitted, be convenient to realize the real-time follow-up of zero deviation.
The reference value quantity at zero point at every turn reading can be set according to accuracy requirement, for example, be generally made as three times, reads respectively the reference value at zero point of confirming and preserving after first three time works on power, obtain respectively three zero point reference value.
B) gather this reference value at original zero point while powering on of torque sensor.
During the each power-up initializing of torque sensor, can read one original zero point reference value, in order to eliminate accidentalia, continuous sampling obtain for eight times eight original zero point reference value, by medium filtering, get four intermediate values, and the mean value of getting these four intermediate values is as reference value at described original zero point; According to the difference of accuracy requirement, continuous sampling number of times also can suitably be adjusted.
C) by step b) original zero point reference value with step the reference value at each in a) at zero point compare.
If the difference of comparing is all in setting range, using step b) the reference value at interim zero point that original zero point, reference value obtained after study as static zero point; Otherwise, step a) in the torque sensor last time confirms after working on power and zero point of preserving, reference value obtained after learning zero point as static state interim zero point reference value.
For example step obtain in a) three zero point reference value, so will be by step b) original zero point reference value respectively with this three zero point reference value compare, therefore also obtain three differences, according to time that three zero points, reference value obtained from the close-by examples to those far off, three corresponding differences should be respectively in: in the scope of 100～300 millivolts, 130～390 millivolts, 150～450 millivolts, using step b) the reference value at interim zero point that original zero point, reference value obtained after study as static zero point.
(2) dynamically revise zero point:
D) in the time window of setting according to predetermined time interval sampling torque sensor signal.
The time window of setting can be 10～60 seconds, for example, be 30 seconds; And predetermined time interval be generally 500～1000 microseconds, 500 microseconds for example.
E) by steps d) in each current sampled value utilize its previous filter value to carry out first-order filtering, obtain the filter value corresponding with each sampled value in time window;
For example, utilize the filter value of N-1 sampled value to carry out first-order filtering to N sampled value, obtain the filter value of N sampled value, this filter value is applied in the first-order filtering of N+1 sampled value again.That is, Y
n=A1 * X
n+ (1-A1) * Y
n-1.Filter factor A1 can be 0.032～0.140, for example, be 0.094.
Wherein:
X
nbe N sampled value;
Y
nfor with N the filter value that sampled value is corresponding;
Y
n-1for with N-1 the filter value that sampled value is corresponding, i.e. previous filter value.
F) by step e) filter value in the same time window that obtains meets limit and weakens integration, be about to each filter value respectively with step c) reference value at interim zero point poor, obtain difference, the difference in 600 millivolts is added up and obtains accumulated value.If accumulated value is positive number, represents to revise to the right, otherwise represent to revise left.
G) utilize described accumulated value to determine the current initial correction value corresponding with this time window with current vehicle speed filter value, utilize the second-order filter value corresponding with the time window of first twice to carry out second-order filter to current initial correction value, obtain the second-order filter value of current initial correction value, if this second-order filter value is greater than 250 millivolts of default safe modified values, using default safe modified value as modified value, otherwise using this second-order filter value as modified value;
Described current vehicle speed filter value obtains in the following way:
By edge, catch and gather vehicle speed sensor signal, according to predetermined time interval obtain speed of a motor vehicle initial value, predetermined time interval be generally 500～1000 milliseconds, for example 500 milliseconds.
Utilize its previous filter value to carry out first-order filtering each current vehicle speed initial value, obtain the filter value corresponding with each initial value;
For example, utilize the filter value of N-1 initial value to carry out first-order filtering to N initial value, obtain the filter value of N initial value, this filter value is applied in the first-order filtering of N+1 initial value again.That is, Y
n=A1 ' * X
n+ (1-A1 ') * Y
n-1.Filter factor A1 ' can be 0.25～0.43, for example, be 0.31.
Wherein:
X
nbe N initial value;
Y
nfor with N the filter value that initial value is corresponding;
Y
n-1for with N-1 the filter value that initial value is corresponding, i.e. previous filter value.
From explaining above, described current vehicle speed filter value the is current time window corresponding filter value of the speed of a motor vehicle while finishing.
While carrying out second-order filter, utilize the filter value of N-2 modified value and the filter value of N-1 modified value to carry out second-order filter to N modified value, obtain the filter value of N modified value, this filter value is applied in the second-order filter of N+1 modified value again.That is, Y
n=A2 * X
n+ B2 * Y
n-1-(A2+B2-1) * Y
n-2.Filter factor A2 is 0.188.Filter factor B2 is 1.664.
Wherein:
X
nbe N initial correction value;
Y
nit is the second-order filter value of N initial correction value;
Y
n-1it is the second-order filter value of N-1 initial correction value;
Y
n-2it is the second-order filter value of N-2 initial correction value.
By step c) reference value at interim zero point be added and obtain modified value at revised zero point at dynamic zero point with the modified value obtaining.
Because the aging of physical construction causes moment of torsion zero deviation, so the trend of zero deviation is the local dull process of analog variation slowly.For this essential characteristic, the present invention is by the static study at zero point of step (1) and the dynamic correction at zero point of step (2), gather and process mass data, filtering disturbance factor effectively, fully guarantee the accuracy of zero correction, can also in user's use procedure, follow the tracks of real-time dynamicly zero deviation.
From step a), each zero correction all can use reference value queue at zero point, in order to guarantee the reliability of reference value queue at zero point, as preferably, in power down keeps, gather power down reference value at zero point constantly, by this power down constantly zero point reference value and step c) reference value at interim zero point poor, if being in (83～250 millivolts) in setting range, difference utilize power down reference value at zero point constantly to upgrade described reference value queue at zero point, be about to power down constantly zero point, reference value was as this end-of-job of torque sensor time reference value at zero point preserve, otherwise utilize step c) reference value at interim zero point upgrade described reference value queue at zero point.
Because torque sensor has certain effective range, generally to there is a torque sensor measuring limit scope, the measurement range of an actual use of torque sensor, and the measurement range of the actual use of torque sensor should be within torque sensor measuring limit scope, and both differences are relocatable spaces, static study at zero point and dynamically revise zero point all should be in this space, otherwise can exceed torque sensor measuring limit scope.
During nominal situation, torque sensor T0 at electric zero point and mechanical zero coincide, the symmetrical output of power-assisted.The higher limit T1 of the actual use of torque sensor, this value is determined when the system matches by real vehicle, motor, tubing string.The higher limit T3 of torque sensor limit range, this value is determined by the linear output area of sensor self.Therefore can calculate the scope of static study at zero point, the actual interval Tr=T1-T0 using, the higher limit Sr=T3-Tr of static study at zero point.After zero correction, be updated to T0 ' zero point of torque sensor, need to guarantee T0 ' < Sr, otherwise cut off immediately power-assisted output, and light trouble light.Dynamic zero point is when driving when revise, T0 ' at zero point after proofreading and correct is finely tuned up and down, this fine setting scope determined jointly by moment of torsion integrated value and current vehicle speed filter value, finally guarantees in certain limit to zero migration real-time follow-up the symmetrical output of power-assisted.
As preferably, learn scope and should be in the measurement range of actual uses of torque sensor and the difference of torque sensor measuring limit scope static zero point, i.e. step c) described interim zero point reference value and the reference value at zero point (initial value before equipment use) given tacit consent to of torque sensor between difference should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope.Otherwise send alerting signal, and cut off the output of electric boosting steering system, guarantee that torque sensor is safe and reliable, avoid distortion work.
As preferably, dynamically zero point, extent of amendment should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, be step g) described modified value, should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope.Otherwise send alerting signal, and cut off the output of electric boosting steering system, guarantee that torque sensor is safe and reliable, avoid distortion work.
As preferably, described static study at zero point scope is greater than extent of amendment at dynamic zero point, and the scope of learning static zero point determines by the effective range of torque sensor, and the safe modified value that dynamically zero point, extent of amendment was set by control algolithm limits.
The present invention also provides a kind of zero-point correction apparatus of torque sensor, comprising:
Modules A, for the front reference value at zero point of confirming after working on power several times and preserving that reads respectively torque sensor, obtain reference value queue at zero point;
Module B, for gathering this reference value at original zero point while powering on of torque sensor;
Module C, for the reference value at original zero point of module B is compared with reference value at each of modules A at zero point;
If the difference of comparing is all in setting range, using the reference value at interim zero point that original zero point, reference value obtained after study as static zero point of module B; Otherwise after the torque sensor last time in modules A is worked on power, confirm and preserve zero point reference value as reference value at interim zero point;
For example step obtain in a) three zero point reference value, so will be by step b) original zero point reference value respectively with this three zero point reference value compare, therefore also obtain three differences, according to time that three zero points, reference value obtained from the close-by examples to those far off, three corresponding differences should be respectively in: in the scope of 100～300 millivolts, 130～390 millivolts, 150～450 millivolts, using step b) the reference value at interim zero point that original zero point, reference value obtained after study as static zero point.
Module D, in the time window of setting according to predetermined time interval sampling torque sensor signal;
The time window of setting can be 10～60 seconds, for example, be 30 seconds; And predetermined time interval be generally 500～1000 microseconds, 500 microseconds for example.
Module E, for utilizing its previous filter value to carry out first-order filtering each current sampled value of module D, obtain the filter value corresponding with each sampled value in time window;
For example, utilize the filter value of N-1 sampled value to carry out first-order filtering to N sampled value, obtain the filter value of N sampled value, this filter value is applied in the first-order filtering of N+1 sampled value again.That is, Y
n=A1 * X
n+ (1-A1) * Y
n-1.Filter factor A1 can be 0.032～0.140, for example, be 0.094.
Wherein:
X
nbe N sampled value;
Y
nfor with N the filter value that sampled value is corresponding;
Y
n-1for with N-1 the filter value that sampled value is corresponding, i.e. previous filter value.
Module F, for the filter value in the same time window that module E is obtained, meet limit and weaken integration, be about to each filter value poor with the reference value at interim zero point of module C respectively, obtain difference, the difference in 600 millivolts is added up and obtains accumulated value;
Module G, for utilizing described accumulated value and current vehicle speed filter value to determine the current initial correction value corresponding with this time window, utilize the second-order filter value corresponding with the time window of first twice to carry out second-order filter to current initial correction value, obtain the second-order filter value of current initial correction value, if this second-order filter value is greater than 250 millivolts of default safe modified values, using default safe modified value as modified value, otherwise using this second-order filter value as modified value;
The reference value at interim zero point of module C is added and obtains modified value at zero point with the modified value obtaining.
Described current vehicle speed filter value obtains in the following way:
By edge, catch and gather vehicle speed sensor signal, according to predetermined time interval obtain speed of a motor vehicle initial value, predetermined time interval be generally 500～1000 milliseconds, for example 500 milliseconds.
Utilize its previous filter value to carry out first-order filtering each current vehicle speed initial value, obtain the filter value corresponding with each initial value;
For example, utilize the filter value of N-1 initial value to carry out first-order filtering to N initial value, obtain the filter value of N initial value, this filter value is applied in the first-order filtering of N+1 initial value again.That is, Y
n=A1 ' * X
n+ (1-A1 ') * Y
n-1.Filter factor A1 ' can be 0.25～0.43, for example, be 0.31.
Wherein:
X
nbe N initial value;
Y
nfor with N the filter value that initial value is corresponding;
Y
n-1for with N-1 the filter value that initial value is corresponding, i.e. previous filter value.
From explaining above, described current vehicle speed filter value the is current time window corresponding filter value of the speed of a motor vehicle while finishing.
While carrying out second-order filter, utilize the filter value of N-2 modified value and the filter value of N-1 modified value to carry out second-order filter to N modified value, obtain the filter value of N modified value, this filter value is applied in the second-order filter of N+1 modified value again.That is, Y
n=A2 * X
n+ B2 * Y
n-1-(A2+B2-1) * Y
n-2.Filter factor A2 is 0.188.Filter factor B2 is 1.664.
Wherein:
X
nbe N initial correction value;
Y
nit is the second-order filter value of N initial correction value;
Y
n-1it is the second-order filter value of N-1 initial correction value;
Y
n-2it is the second-order filter value of N-2 initial correction value.
As preferably, after working on power, first three time of reading respectively torque sensor in modules A confirm and the reference value at zero point of preserving obtains reference value queue at zero point.
As preferably, in module B continuous sampling obtain for eight times eight original zero point reference value, by medium filtering, get four intermediate values, and the mean value of getting these four intermediate values is as reference value at described original zero point.
As preferably, the time window described in module D is 10～60 seconds, and predetermined time is spaced apart 500～1000 microseconds.
As preferably, be provided with reference value queue at zero point update module, for keeping in power down, gather power down reference value at zero point constantly, this power down reference value at zero point constantly is poor with the reference value at interim zero point of module C, if difference is in the reference value queue at zero point described in the reference value renewal at zero point that (83～250 millivolts) in setting range utilize the power down moment; Otherwise utilize the described reference value queue at zero point of interim reference value renewal at zero point of module C.
In described module C, difference between interim zero point, reference value was given tacit consent to torque sensor zero point reference value should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, otherwise send alerting signal, and cut off the output of electric boosting steering system.
In described module G, modified value, should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, otherwise send alerting signal, and cut off the output of electric boosting steering system.
Zero correction method of the present invention is divided into static study at zero point and dynamically revises zero point, and according to certain sequential and the mutual verification confirmation of algorithm, with filtering disturbance, guarantees the accuracy of controlling; Can in user's use procedure, follow the tracks of real-time dynamicly zero deviation.The present invention is limited to control algolithm in the limit range of torque sensor, strictly guarantee the security of zero correction, guarantee in user's driving procedure, steering is symmetrical equilibrium all the time, even if exceed certain zero correction scope, controller also can cut off power-assisted output immediately, lights trouble light, gives to greatest extent user's driver comfort and security; Make electric boosting steering system reach optimum matching, extend the service time of product.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of Zero correction method of the present invention.
Embodiment
The zero-point correction apparatus of a kind of torque sensor of the present invention, comprising:
Modules A, for the front reference value at zero point of confirming after working on power several times and preserving that reads respectively torque sensor, obtain reference value queue at zero point;
Module B, for gathering this reference value at original zero point while powering on of torque sensor;
Module C, for the reference value at original zero point of module B is compared with reference value at each of modules A at zero point;
If the difference of comparing is all in setting range, using the reference value at interim zero point that original zero point, reference value obtained after study as static zero point of module B; Otherwise after the torque sensor last time in modules A is worked on power, confirm and preserve zero point reference value as reference value at interim zero point;
Module D, in the time window of setting according to predetermined time interval sampling torque sensor signal;
Module E, for utilizing its previous filter value to carry out first-order filtering each current sampled value of module D, obtain the filter value corresponding with each sampled value in time window;
Module F, for the filter value in the same time window that module E is obtained, meet limit and weaken integration, be about to each filter value poor with the reference value at interim zero point of module C respectively, obtain difference, the difference in setting range is added up and obtains accumulated value;
Module G, for utilizing described accumulated value and current vehicle speed filter value to determine the current initial correction value corresponding with this time window, utilize the second-order filter value corresponding with the time window of first twice to carry out second-order filter to current initial correction value, obtain the second-order filter value of current initial correction value, if this second-order filter value is greater than default safe modified value, using default safe modified value as modified value, otherwise using this second-order filter value as modified value;
The reference value at interim zero point of module C is added and obtains modified value at zero point with the modified value obtaining.
Also be provided with reference value queue at zero point update module, for keeping in power down, gather power down reference value at zero point constantly, this power down reference value at zero point constantly is poor with the reference value at interim zero point of module C, if difference is in the reference value queue at zero point described in the reference value renewal at zero point that utilizes the power down moment in setting range; Otherwise utilize the described reference value queue at zero point of interim reference value renewal at zero point of module C.
Referring to Fig. 1, the Zero correction method of a kind of torque sensor of the present invention, can utilize zero-point correction apparatus of the present invention to realize, the corresponding step of modules A～G a)～g), in corresponding module, carry out respectively following steps:
(1) static study at zero point:
A) controller is confirmed after working on power several times before reading from nonvolatile memory and the reference value at zero point of preserving obtains reference value queue at zero point zero_cfm_base[n].
Get n=3 herein, queue adopts FIFO form, be zero_cfm_base[0] be the reference value at zero point of preserving when power down keeps after front one action, zero_cfm_base[1] be again the reference value at zero point of preserving when power down keeps after front one action, zero_cfm_base[2] be the reference value at zero point of preserving when power down keeps after more front one action.
B) gather this reference value zero_raw_base1 at original zero point while powering on of torque sensor.
Continuous sampling torque sensor signal 8 times, obtains 8 crude sampling value t0, t1, t2, t3, t4, t5, t6, t7, obtains 4 intermediate values, then by mean filter, obtain the mean value t of 4 intermediate values by medium filtering, gets zero_raw_base1=t.
C) by step b) original zero point reference value zero_raw_base1 with step a) in each at zero point reference value compare, use zero_cfm_base[2], zero_cfm_base[1], zero_cfm_base[0] zero_raw_base1 is carried out to the differentiation verification of movement tendency.
As skew to the right,
Work as zero_cfm_base[0] < zero_raw_base1, and meet zero_raw_base1-zero_cfm_base[0] < RR * K0;
Work as zero_cfm_base[0] > zero_raw_base1, and meet zero_cfm_base[0]-zero_raw_base1 < RL * K0.
Work as zero_cfm_base[1] < zero_raw_base1, and meet zero_raw_base1-zero_cfm_base[1] < RR * K1;
Work as zero_cfm_base[1] > zero_raw_base1, and meet zero_cfm_base[1]-zero_raw_base1 < RL * K1.
Work as zero_cfm_base[2] < zero_raw_base1, and meet zero_raw_base1-zero_cfm_base[2] < RR * K2;
Work as zero_cfm_base[2] > zero_raw_base1, and meet zero_cfm_base[2]-zero_raw_base1 < RL * K2.
As skew left,
Work as zero_cfm_base[0] > zero_raw_base1, and meet zero_cfm_base[0]-zero_raw_base1 < LL * K0;
Work as zero_cfm_base[0] < zero_raw_base1, and meet zero_raw_base1-zero_cfm_base[0] < LR * K0.
Work as zero_cfm_base[1] > zero_raw_base1, and meet zero_cfm_base[1]-zero_raw_base1 < LL * K1;
Work as zero_cfm_base[1] < zero_raw_base1, and meet zero_raw_base1-zero_cfm_base[1] < LR * K1.
Work as zero_cfm_base[2] > zero_raw_base1, and meet zero_cfm_base[2]-zero_raw_base1 < LL * K2;
Work as zero_cfm_base[2] < zero_raw_base1, and meet zero_raw_base1-zero_cfm_base[2] < LR * K2.
RR=LL=300 millivolt, RL=LR=100 millivolt.Correlation coefficient K0=1, K1=1.3, K2=1.5.Consider symmetrical property, RR=LL, RL=LR.RR represents that with LL movement tendency is consistent, and RL and LR represent movement tendency fluctuation, thus RR > RL, LL > LR, the limit value that trend is consistent is large, and the limit value of trend fluctuation is little.Because zero migration under actual condition is a local dull process of analog variation slowly, and in conjunction with the preservation sequential of reference value at zero point, so K0 < K1 < K2.
When zero_raw_base1 meets all criterions, interim zero point reference value zero_temp_base=zero_raw_base1, otherwise interim zero point reference value zero_temp_base=zero_cfm_base[0].
(2) dynamically revise zero point:
D) in the time window of setting according to predetermined time interval sampling torque sensor signal, the time window of setting is 30 seconds; And predetermined sampling time interval is generally 500 microseconds.
E) by steps d) in each current sampled value t_atd, filter value t_1filtered carried out first-order filtering with last time, obtain filter value t_1filt=A1 * t_atd+ (1-A1) * t_1filtered, filter factor A1 can be 0.032～0.140, for example, be 0.094.Filter factor has determined that the cutoff frequency of sampled value and phase place lag behind.A1 value is large, and torque sensor signal response is sensitive, but is easily subject to the disturbance of high frequency noise, poor stability; A1 value is little, torque sensor signal response lag, but Ability of Resisting Disturbance is strong, good stability.
F) in each time window of 30 seconds, a plurality of filter values that calculate are above met to limit and weaken integration.
As dextrad, difference t_delt=filter value t_1filt-reference value zero_temp_base at interim zero point is positive number, if | difference t_delt| > C, this value is given up, otherwise difference t_delt is cumulative;
As left-hand, difference t_delt=filter value t_1filt-reference value zero_temp_base at interim zero point is negative, if | difference t_delt| > C, this value is given up, otherwise difference t_delt is cumulative.
C can be 500～800 millivolts, for example, be 600 millivolts.
If the accumulated value of final difference is positive number, represents that (> zero_temp_base) revises to the right, otherwise represent that (< zero_temp_base) revises left.
G) according to step f) accumulated value obtaining and the filter value of current vehicle speed definite current initial correction value t_tab corresponding with this time window in table 1.
Current vehicle speed filter value obtains in the following way:
By edge, catch and gather vehicle speed sensor signal, according to predetermined time interval obtain speed of a motor vehicle initial value, predetermined time interval be generally 500 milliseconds.
By each speed of a motor vehicle initial value v_dat, filter value v_1filtered carried out first-order filtering with last time, obtained filter value v_1filt=A1 ' * v_dat+ (1-A1 ') * v_1filtered, and filter factor A1 ' can be 0.25～0.43, for example, be 0.31.Filter factor has determined that the cutoff frequency of vehicle speed value and phase place lag behind.A1 ' value is large, and vehicle speed sensor signal response is sensitive, but is easily subject to the disturbance of high frequency noise, poor stability; A1 ' value is little, and vehicle speed sensor signal response lags behind, but Ability of Resisting Disturbance is strong, good stability.
From explaining above, current vehicle speed filter value the is current time window corresponding filter value of the speed of a motor vehicle while finishing.
Table 1
In table 1:
X axle represents the accumulated value that the moment of torsion accumulated value (being corresponding step f) of certain direction obtains); Reflection zero migration trend, along with the increase of accumulated value, the modified value of z axle output also can increase;
Y axle represents current vehicle speed filter value (kilometer per hour), and along with the increase of the speed of a motor vehicle, required power steering can reduce rapidly, and for guaranteeing stability and the security of driving, the modified value of z axle output also can reduce.
Z axle is initial correction value (millivolt), i.e. the current initial correction value t_tab corresponding with this time window.If accumulated value step f) obtaining and current vehicle speed filter value do not have accurately corresponding numerical value in table 1, according to the numerical value change trend of x-axis, y-axis and z-axis, make linear interpolation digital simulation and draw empirical curve, current initial correction value t_tab utilizes this empirical curve to determine.
Second-order filter value t_2filtered0, the t_2filteredl of initial correction value t_tab definite in table 1 and front twice time window are carried out to second-order filter, obtain second-order filter value t_2filt=A2 * t_tab+B2 * t_2filtered0-(A2+B2-1) * t_2filteredl, filter factor A2 is 0.188.Filter factor B2 is 1.664.Filter factor has determined that the cutoff frequency of modified value and phase place lag behind, and according to torque sensor family curve, motor characteristic curve, mechanical tubing string ratio of gear, driving dynamic model, power-assisted output, controls the system matches such as model because usually determining.Because zero migration is a local dull process of analog variation slowly, and more will guarantee the security of driving a vehicle under operating mode, so adopt, Ability of Resisting Disturbance is strong, the second-order filter algorithm of good stability.
By dynamic zero point expurgated bound fix on safe scope, if second-order filter value t_2filt > D, modified value zero_offset=D, otherwise modified value zero_offset=second-order filter value t_2filt.
D can be 200～300 millivolts, for example, be 250 millivolts.
Final controller in conjunction with before the correction direction that obtains, obtain modified value at revised zero point at dynamic zero point=interim reference value zero_temp_base+ modified value zero_offset at zero point.
Controller gathers power down reference value zero_raw_base2 at zero point constantly in this power down keeps.
This power down reference value zero_raw_base2 at zero point is constantly also according to step b) mode obtain, be continuous sampling torque sensor signal 8 times, obtain 8 crude sampling values, by medium filtering, obtain 4 intermediate values, by mean filter, obtain again the mean value of 4 intermediate values, take down electricity reference value zero_raw_base2=mean value at zero point constantly.
Controller uses interim zero point reference value zero_temp_base to carry out the differentiation verification of movement tendency to power down reference value zero_raw_base2 at zero point constantly.
With step c) adopt identical algorithm,
As skew to the right,
As zero_temp_base < zero_raw_base2, and meet zero_raw_base2-zero_temp_base < RR * K;
As zero_temp_base > zero_raw_base2, and meet zero_temp_base-zero_raw_base2 < RL * K.
As skew left,
As zero_temp_base > zero_raw_base2, and meet zero_temp_base-zero_raw_base2 < LL * K;
As zero_temp_base < zero_raw_base2, and meet zero_raw_base2-zero_temp_base < LR * K.
RR, LL, RL, LR is step c) in parameter.Correlation coefficient K=0.83.The corresponding the last movement tendency of K is differentiated, so K is less than step c) in front movement tendency several times K0, the K1, the K2 that differentiate, i.e. K < K0 < K1 < K2.
When zero_raw_base2 meets criterion, upgrade as follows reference value queue at zero point:
zero_cfm_base[2]＝zero_cfm_base[1]、
zero_cfm_base[1]＝zero_cfm_base[0]、
zero_cfm_base[0]＝zero_raw_base2。
Otherwise upgrade as follows reference value queue at zero point:
zero_cfm_base[2]＝zero_cfm_base[1]、
zero_cfm_base[1]＝zero_cfm_base[0]、
zero_cfm_base[0]＝zero_temp_base。
The reference value queue at zero point of renewal is saved in nonvolatile memory, has guaranteed reliability, the transitivity of reference value queue at zero point.
In Zero correction method of the present invention, also requiring to learn scope static zero point should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, dynamically zero point, extent of amendment should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, and static study at zero point scope is greater than extent of amendment at dynamic zero point, otherwise send alerting signal, and cut off the output of electric boosting steering system, guarantee that torque sensor is safe and reliable, avoid distortion work.
Claims (10)
1. a Zero correction method for torque sensor, is characterized in that, comprises the steps:
(1) static study at zero point:
A) the front reference value at zero point of confirming after working on power several times and preserving that reads respectively torque sensor obtains reference value queue at zero point;
B) gather this reference value at original zero point while powering on of torque sensor;
C) reference value at original zero point of step b) is compared with the reference value at each in step a) at zero point;
If the difference of comparing is all in setting range, using the reference value at interim zero point that original zero point, reference value obtained after study as static zero point of step b); Otherwise the reference value at interim zero point that zero point, reference value obtained after study as static zero point of confirming after the torque sensor last time in step a) is worked on power and preserving;
(2) dynamically revise zero point:
D) in the time window of setting according to predetermined time interval sampling torque sensor signal;
E) utilize its previous filter value to carry out first-order filtering each the current sampled value in step d), obtain the filter value corresponding with each sampled value in time window;
F) filter value in same time window step e) being obtained is met limit and is weakened integration, is about to each filter value poor with the reference value at interim zero point of step c) respectively, obtains difference, and the difference in setting range is added up and obtains accumulated value;
G) utilize described accumulated value to determine the current initial correction value corresponding with this time window with current vehicle speed filter value, utilize the second-order filter value corresponding with the time window of first twice to carry out second-order filter to current initial correction value, obtain the second-order filter value of current initial correction value, if this second-order filter value is greater than default safe modified value, using default safe modified value as modified value, otherwise using this second-order filter value as modified value; Described current vehicle speed filter value the is current time window corresponding filter value of the speed of a motor vehicle while finishing;
The reference value at interim zero point of step c) is added and obtains modified value at revised zero point at dynamic zero point with the modified value obtaining.
2. the Zero correction method of torque sensor as claimed in claim 1, is characterized in that, confirms and the reference value at zero point of preserving obtains reference value queue at zero point in step a) after first three that reads respectively torque sensor time works on power.
3. the Zero correction method of torque sensor as claimed in claim 2, it is characterized in that, in step b) continuous sampling obtain for eight times eight original zero point reference value, by medium filtering, get four intermediate values, and the mean value of getting these four intermediate values is as reference value at described original zero point.
4. the Zero correction method of torque sensor as claimed in claim 3, is characterized in that, the time window described in step d) is 10～60 seconds, and predetermined time is spaced apart 500～1000 microseconds.
5. the Zero correction method of the torque sensor as described in claim 1～4 any one, it is characterized in that, in power down keeps, gather power down reference value at zero point constantly, this power down reference value at zero point constantly is poor with the reference value at interim zero point of step c), if difference is in the reference value queue at zero point described in the reference value renewal at zero point that utilizes the power down moment in setting range; Otherwise utilize the described reference value queue at zero point of interim reference value renewal at zero point of step c).
6. the Zero correction method of torque sensor as claimed in claim 5, it is characterized in that, described in step c) interim zero point reference value and zero point of torque sensor acquiescence the difference between reference value should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, otherwise send alerting signal, and cut off the output of electric boosting steering system.
7. the Zero correction method of torque sensor as claimed in claim 6, it is characterized in that, modified value described in step g) should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, otherwise send alerting signal, and cut off the output of electric boosting steering system.
8. a zero-point correction apparatus for torque sensor, is characterized in that, comprising:
Modules A, for the front reference value at zero point of confirming after working on power several times and preserving that reads respectively torque sensor, obtain reference value queue at zero point;
Module B, for gathering this reference value at original zero point while powering on of torque sensor;
Module C, for the reference value at original zero point of module B is compared with reference value at each of modules A at zero point;
If the difference of comparing is all in setting range, using the reference value at interim zero point that original zero point, reference value obtained after study as static zero point of module B; Otherwise after the torque sensor last time in modules A is worked on power, confirm and preserve zero point reference value as reference value at interim zero point;
Module D, in the time window of setting according to predetermined time interval sampling torque sensor signal;
Module E, for utilizing its previous filter value to carry out first-order filtering each current sampled value of module D, obtain the filter value corresponding with each sampled value in time window;
Module F, for the filter value in the same time window that module E is obtained, meet limit and weaken integration, be about to each filter value poor with the reference value at interim zero point of module C respectively, obtain difference, the difference in setting range is added up and obtains accumulated value;
Module G, for utilizing described accumulated value and current vehicle speed filter value to determine the current initial correction value corresponding with this time window, utilize the second-order filter value corresponding with the time window of first twice to carry out second-order filter to current initial correction value, obtain the second-order filter value of current initial correction value, if this second-order filter value is greater than default safe modified value, using default safe modified value as modified value, otherwise using this second-order filter value as modified value; Described current vehicle speed filter value the is current time window corresponding filter value of the speed of a motor vehicle while finishing;
The reference value at interim zero point of module C is added and obtains modified value at zero point with the modified value obtaining.
9. the zero-point correction apparatus of torque sensor as claimed in claim 8, it is characterized in that, be provided with reference value queue at zero point update module, for keeping in power down, gather power down reference value at zero point constantly, this power down reference value at zero point constantly is poor with the reference value at interim zero point of module C, if difference is in the reference value queue at zero point described in the reference value renewal at zero point that utilizes the power down moment in setting range; Otherwise utilize the described reference value queue at zero point of interim reference value renewal at zero point of module C.
10. the zero-point correction apparatus of torque sensor as claimed in claim 9, it is characterized in that, in described module C, difference between interim zero point, reference value was given tacit consent to torque sensor zero point reference value should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, otherwise send alerting signal, and cut off the output of electric boosting steering system;
In described module G, modified value should be in the measurement range of the actual use of torque sensor and the difference of torque sensor measuring limit scope, otherwise sends alerting signal, and cuts off the output of electric boosting steering system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110449600.2A CN102564689B (en) | 2011-12-29 | 2011-12-29 | Zero correction method and device of torque sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110449600.2A CN102564689B (en) | 2011-12-29 | 2011-12-29 | Zero correction method and device of torque sensor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102564689A CN102564689A (en) | 2012-07-11 |
| CN102564689B true CN102564689B (en) | 2014-02-05 |
Family
ID=46410703
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110449600.2A Expired - Fee Related CN102564689B (en) | 2011-12-29 | 2011-12-29 | Zero correction method and device of torque sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102564689B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8825279B2 (en) * | 2012-09-11 | 2014-09-02 | Shimano Inc. | Bicycle power sensing apparatus |
| CN103808452B (en) * | 2012-11-15 | 2016-09-14 | 台达电子工业股份有限公司 | The zero adjustment method of electric power-assisted steering apparatus and balancing controls thereof |
| CN103085864B (en) * | 2013-01-25 | 2015-07-29 | 株洲易力达机电有限公司 | A kind of electric boosting steering system torque sensor drift compensating method |
| CN103646591B (en) * | 2013-12-16 | 2016-02-10 | 北京经纬恒润科技有限公司 | A kind of sensor zero point learning method and system |
| CN103743508B (en) * | 2013-12-27 | 2015-12-02 | 中国科学院苏州生物医学工程技术研究所 | Torsional moment test macro and torque sensor device thereof |
| CN103728904B (en) * | 2014-01-06 | 2016-08-17 | 金东纸业（江苏）股份有限公司 | Encoder correction system and method |
| DE102015216339B3 (en) * | 2015-08-26 | 2016-09-08 | Continental Automotive Gmbh | Method for calibrating a arranged in the drive train of a motor vehicle torque sensor |
| CN106996341B (en) * | 2017-03-27 | 2019-08-23 | 宁波吉利汽车研究开发有限公司 | A kind of diagnostic method of broad domain oxygen sensor response failure |
| JP2019117103A (en) * | 2017-12-27 | 2019-07-18 | アズビル株式会社 | Correction method of zero point error in torque sensor |
| CN108489589A (en) * | 2018-03-15 | 2018-09-04 | 天津光电丰泰科技有限公司 | A kind of drift correction method based on weighing sensor |
| CN108839559B (en) * | 2018-06-21 | 2020-07-17 | 北京新能源汽车股份有限公司 | Self-adaptive learning method and device for zero position of accelerator pedal and vehicle |
| CN111189577B (en) * | 2020-01-16 | 2022-01-07 | 腾讯科技（深圳）有限公司 | Sensor calibration and data measurement method, device, equipment and storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100419397C (en) * | 2005-06-14 | 2008-09-17 | 探微科技股份有限公司 | Method for correcting zero offset voltage of pressure transducer |
| CN102087157B (en) * | 2010-11-16 | 2013-01-02 | 吉林大学 | Double-lever force standard machine and calibration and demarcation method thereof |
-
2011
- 2011-12-29 CN CN201110449600.2A patent/CN102564689B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN102564689A (en) | 2012-07-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102564689B (en) | Zero correction method and device of torque sensor | |
| WO2019242718A1 (en) | Vehicle lane change control method and device | |
| CN101842279B (en) | Steering control device and vehicle steering device using the same | |
| CN109799701B (en) | Industrial robot vibration suppression method | |
| US6803736B1 (en) | Control system which carries out the model-supported safety monitoring of an electronically regulated controller in the motor vehicle | |
| CN101746412B (en) | Detection of driver intervention during a torque overlay operation in an electric power steering system | |
| CN103003133B (en) | The compensation method of bearing circle torsional oscillation and device in steering swivel system | |
| CN104122531B (en) | The method of self-adaptive processing radar antenna position oscillation | |
| Mammar et al. | Two-degree-of-freedom formulation of vehicle handling improvement by active steering | |
| US20040257022A1 (en) | Method and system for multiple servo motor control | |
| CN101332834A (en) | Control apparatus for electric power steering system | |
| WO2008038117A3 (en) | Vehicle steering system and method of controlling the same | |
| CN102006011A (en) | Controller of electric motor having function of estimating inertia and friction simultaneously | |
| WO2007141578A3 (en) | Power tool control systems | |
| CN108340966A (en) | Position self-learning method in electric boosting steering system and its steering wheel angle | |
| EP2113444A3 (en) | Systems and methods involving torque disturbance rejection | |
| CN101017367A (en) | Controller structure | |
| CA2886126A1 (en) | Ultrasonic motor control system and method | |
| JP6900394B2 (en) | How to operate the auxiliary steering with indirect reproduction of steering wheel torque information | |
| CN105429550A (en) | Current sampling error correction method and current sampling error correction system | |
| CN104590253A (en) | Yaw angular velocity control method for four-wheel independent driving electric vehicle | |
| Byrne et al. | Experimental results in robust lateral control of highway vehicles | |
| CN104802611A (en) | Temperature control method based on navigation equipment and navigation equipment | |
| CN105388853A (en) | numerical control device | |
| CN103157851B (en) | Flying shear control system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140205 Termination date: 20141229 |
|
| EXPY | Termination of patent right or utility model |