Procedure and system for detecting collisions and controlling safety

Claim 1

A method of detecting collisions and controlling safety devices for vehicle passengers in which the vehicle decelerations are continuously measured and in which the measured signals are converted into a numerical signal which represents the deceleration of the vehicle with said numerical signal being processed and compared with predetermined values so that the release of the safety devices is controlled as a function of the result of the comparison, comprising the following steps carried out in cycles with each cycle having a constant duration: calculating a weighted acceleration value based upon said numerical deceleration value for a first cycle and at least one of three constants; producing an mean value of the weighted acceleration value; comparing said mean value with at least a first and a second predetermined value; choosing one of said constants to calculate the weighted acceleration value during the following cycle as a function of the result of said comparison; calculating a correcting term value each time said mean value exceeds successively said first predetermined value and said second predetermined value wherein said correcting term value is a function of the time between the successive excessions of said two predetermined values; retaining said calculated correcting term value as said correcting term as long as said mean value is greater than or equal to said first predetermined value but is less than the value of said second predetermined value; assigning to said correcting term a value of zero when said mean value decreases below said first predetermined value; calculating an algorithm having a value equal to the sum of said mean value and said correcting term; comparing said algorithm value to at least one threshold value; and releasing at least one of said safety devices when said algorithm value is greater than or equal to said at least one threshold value.

Claim 2

A method of detecting collisions and controlling safety devices for vehicle passengers in which the vehicle decelerations are continuously measured and in which the measured signals are converted into a numerical signal which represents the deceleration of the vehicle with said numerical signal being processed and compared with predetermined values so that the release of the safety devices is controlled as a function of the result of the comparison, comprising the following steps carried out in cycles with each cycle having a constant duration: calculating a weighted acceleration value based upon said numerical deceleration value for a first cycle and at least one of three constants; producing an mean value of the weighted acceleration value; comparing said mean value with at least a first and a second predetermined value; choosing one of said constants to calculate the weighted acceleration value during the following cycle as a function of the result of said comparison; calculating a correcting term value each time said mean value exceeds successively said first predetermined value and said second predetermined value wherein said correcting term value is a function of the time between the successive excessions of said two predetermined values; retaining said calculated correcting term value as said correcting term as long as said mean value is greater than or equal to said first predetermined value but is less than the value of said second predetermined value; assigning to said correcting term a value of zero when said mean value decreases below said first predetermined value; calculating an algorithm having a value equal to the sum of said mean value and said correcting term; comparing said algorithm value to at least one threshold value; and releasing at least one of said safety devices when said algorithm value is greater than or equal to said at least one threshold value.

Claim 4

The method according to one of the claims 1 and 2, wherein said mean value is calculated according to the following formula: S.sub.(n+1) =S.sub.(n) +γp-(S.sub.(n) /α) in which S.sub.(n+1) represents the actual value of the mean value, S.sub.(n) represents the value of the exponential average calculated during the preceding cycle, (γp) represents the value of the weighted acceleration which will be calculated at the beginning of the cycle in progress where α is a constant.

Claim 5

The method according to one of the claims 1 and 2, wherein the value of the correcting term is calculated by means of a counter which is set off at a predetermined number (Co) when said mean value S.sub.(n+1) exceeds the first predetermined value (S 1 ) and which is decreased in each cycle, the counting of the counter being interrupted when said mean value (S n+1 ) exceeds the second predetermined value (S 2 ) after having exceeded the first (S 1 ) or when the counter contents reach a zero value, the correcting term being a function of the contents of the counter at the moment of said interruption multiplied by a coefficient (2 1 ).

Claim 6

A system for detecting collisions and controlling safety devices comprising: at least one capacitor deceleration sensor, a forming circuit for the output signal of the sensor, a numerical-analog converter, a numerical calculator and at least one actuator controlled by the numerical calculator, wherein said forming circuit (2) comprises an electronic commutator to polarize said sensor (1) at a zero voltage for a predetermined period, then rendered non-conductive by the calculator (5) to charge the sensor at a predetermined polarization, the calculator (5) being adapted to compare the charge level of the sensor with a predetermined charge level.

Claim 7

A system for detecting collisions and controlling safety devices comprising: at least one capacitor deceleration sensor, a forming circuit for the output signal of the sensor, a numerical-analog converter, a numerical calculator and at least one actuator controlled by the numerical calculator, wherein said forming circuit (2) comprises an electronic commutator to polarize said sensor (1) at a zero voltage for a predetermined period, then rendered non-conductive by the calculator (5) to charge the sensor at a predetermined polarization, the calculator (5) being adapted to compare the charge level of the sensor with a predetermined charge level.