US3588002A - Adaptive missile guidance systems - Google Patents

Abstract

D R A W I N G
A MISSILE GUIDANCE STATION INCLUDING A RESOLVER FOR MODIFYING PITCH AND YAW DEMAND SIGNALS IN ACCORDANCE WITH AN ASSUMED MISSILE ROLL ANGLE, WHICH SIGNALS ARE TRANSMITTED TO THE MISSILE. MODELS OF THE PITCH AND YAW GUIDANCE LOOPS AT THE STATION RECEIVE SIGNALS REPRESENTING REQUIRED MISSILE PITCH AND YAW ANGLES. THE PERFORMANCE OF THESE MODELS IS COMPARED WITH THAT OF THE MISSILE AND THE DIFFERENCES IN PITCH AND YAW PERFORMANCE ARE USED TO ADJUST THE ANGLE OF THE RESOLVER.

Description

United States Patent [72] Invento Al r J- hi e 3,360,2[4 12/1967 Stcherbatcheff 244 114 56. -y L orlh. 3,372.88) 3/1968 Menke 244/314 I N g g g' FOREIGN PATENTS P 763,029 7/1967 Canada 244/114 [22] FM 1967 1 198 209 8/1965 0 244/3 14 [45] Patented June 28, ermany Primary Examiner-Verlin R. Pendegrass At10rneySughrue. Rothwell, Mion, Zinn and MacPcak [54] ADAPTIVE MISSILE GUIDANCE SYSTEMS 5 Claims, 2 Drawing Figs.

[52] US. Cl. 244/114 ABSTRACT: A missile guidance station including a sob-er [5]] F42. for modifying pitch and yaw demand signals in accordance [50] 244/3.l1- with an assumed missile r0" angle hi signals are "ans, mitted to the missile. Models of the pitch and yaw guidance loops at the station receive signals representing required mis- [56] Belem cm sile pitch and yaw angles. The performance of these models is UNITED STATES PATENTS compared with that of the missile and the difi'erences in pitch 3,332,641 7/1967 Bezerie 244l3.l2 and yaw performance are used to adjust the angle of the 3,351,847 ll/l967 Jansson etal 244/3.l2X resolver.

F FIKrkZQJAh/iri'? PATENTEDJ1111281911 SHEET 1 [IF 2 -1 mQ Q W n 1 w m m 1 T :38 aum 38 k i Q m $3, m

m fi v W Q 1 1., ,1 WWQQFWQM 35$ 1 I SHEET 2 OF 2 PATENTEU JUN28 m ADAPTIVE MISSILE GUIDANCE SYSTEMS A freely rolling missile requires some method of translating the ground command signals into signals referred to the missile axes. This is usually accomplished by means of a gyro reference and a roll resolver in the missile. but the resolver axes may not be in phase with the axes of the missile. in which case there is a phasing error" between the required and the demanded components of missile acceleration. The purpose of the present invention is to remove the need for the resolver in the missile. or to overcome the phasing error if such a resolver is provided.

In order that the invention may be better understood, a typical guidance system of the kind to which the invention is applicable will first be described with reference to FIG. I of the accompanying drawings. after which an example of a system embodying the invention will be described with reference to FIG. 2.

In a typical guidance system of the kind to which the invention is applicable. a tracking device provides signals which correspond to the angular positions of the target in azimuth and elevation and also signals X, and Y which correspond to the angular errors between the target and missile in azimuth and elevation. The error signals then pass through suitable shaping networks W and are sent from a transmitter T to a receiver U in the missile as demands for missile accelerations in azimuth and elevation. In this example, each shaping network W contains a multiplication by missile range to convert angular error into distance off the sightline. a proportionality factor to convert distance off the sightline into a demand for missile acceleration and an electrical lead network to compensate for the lags inherent in the missile kinematics. The network output is a demand for missile acceleration in pitch or yaw, for example in the form of a demand for a predetermined number of meters/sec of missile acceleration per meter of error. (or per milliradian of angular error. in the absence of a range multiplication in the shaping network). A computed bias may be introduced into the shaping network to allow for target velocity. If the missile has no autopilot, the demand may be in the form of: radians of fin angle per meter oferror.

If these signals were fed directly to the pitch and yaw autopilots they would not produce the desired response because the missile would have rolled through some angle. 9., say. about its longitudinal axis. For example. if 6,, was 90 then a demand from the tracker to the pitch autopilot for a movement in elevation would result in the movement being executed in the azimuth direction. To overcome this. the angle through which the missile has rolled since launch is measured by means of a gyroscope and this angle. 8 is fed to a roll resolver 9 in the missile. The resolver unit then translates signals referred to the tracker axes into signals referred to a set of coordinate axes which have been rotated through an angle 9,; with respect to the orientation of the missile axes at launch (i.e. into missile axes if there is no phasing error and 8 is equal to 9...). These signals, when their components have been suitably combined in the summing amplifiers l0 and II. are applied to the pitch and yaw autopilots.

The missile accelerations in the pitch and yaw planes are observed by the tracker as motion in the elevation and azimuth planes. is. they are referred to its own system of coordinates. There is thus an implied resolution from missile to tracker axes and this is represented by the implied resolver unit 12. The resolver 12. with its summing amplifiers I3 and 14, the double integrators l5 and the reciprocal missile range units 16, together represent (in simplified form) the missile kinematics" 17 which convert missile accelerations into angular displacements seen by the tracker. The contents of the box 17 are thus imaginary components. The amplifiers I8 and I9 represent a differencing operation carried out in the tracker.

A major disadvantage ofthe system described lien in the fact that gvroscnpes to measure roll angle are expensive and liable to inaccuracies (such Ill gyro drift 1.

According to the present invention. the roll resolver is located at the flied station and the angle of resolution of the resolver is adjusted by means responsive to the errors between the actual and required angular positions of the missile in pitch and yaw, as seen from the ground station and also to a circuit for determining the polarity of the error of resolution angle; the adjustment is in such a sense as to decrease the errors in pitch and ysw. In the preferred form. models of the pitch and yaw guidance loops are located at the fixed station. and means are provided for comparing the observed output of the actual guidance loop. that is to say. the movement in elevation and azimuth of the missile. with the output of the fixed models, the diiference between these outputs constituting actuating signals for adjusting the said resolver at the fixed station. In this way the fixed station resolver is kept substantially in step with the rotation of the missile in roll and as a consequence the signals transmitted from the tracker station represent pitch and yaw demands to the missile. whatever the angle through which it has rolled since launching.

A block diagram of this self-adaptive system will now be described with reference to FIG. 2. The elements of the diagram of FIG. I are present in the middle portion of FIG. 2. but the latter FIG. additionally contains representations of the electrical models of the pitch and yaw guidance loops. which are sited in or near the ground tracker, and the circuits interconnecting these loops and the tracker. The models. which are shown at the top and bottom of FIG. 2. represent the ideal guidance loops which would operate if there were no errors due to the missile roll position and they can be greatly simplified (for example. the autopilot in each model loop could be neglected). In the central portion of the diagram, the resolver 20 with its summing amplifiers 21 and 22. is now located on the ground near the tracker. in the forward paths of the guidance loops. This resolver is continuously rotated by the self-adaptive system in such a way that 6,, is approximately equal to 6,, (the missile roll angle) at all timesv A list of symbols to be used in the following description will now be given. together with the parameters represented by these symbols.

X Y, Command input signals to guidance loops (target angular positions) a... 1, Adaptive loop actuating signals.

X Y; Guidance loop error signals.

X,,. Y Loop error signals in fixed models.

G Gain factor in adaptive loop.

d p(- Laplact. operator.

The command input signals X and Y (corresponding to the target angular position relative to the tracker) are fed to the models. and the guidance loop error signals X; and Y; (corresponding to the displacements of the missile image from the center of the tracking head. which is pointed at the target at all times) are fed to the missile via suitable shaping networks W and roll resolutions obtained by means of the resolver 20.

The "feedback" due to the missile kinematics takes the form of the displacements of the missile image from the center of the tracking head and the signals X and Y; are thus electronics in origin. The same target angle signals X. and Y. are applied to the pitch and yaw models. each of which contains a shaping network W, an autopilot simulator and a double integrator 29. These models have feedback loops leading to input differencing amplifiers 30 and 31 and the signals X, and Y from the latter are thus comparable with those from the amplifiers I8 and 19. These pairs of pitch and yaw signals from the amplifiers I8 and 30 (pitch) and 19 and 31 (yaw) are in fact compared in differencing amplifiers 32 and 33. the outputs of which represent the difference in response between the actual missile and the models due to inequality of 9,, and 9, These output signals (1, and 1, constitute the adaptive loop actuating signals.

These signals must then be multiplied by parameter sensitivlty signals which correspond to their rates of change. respectively. with the resolver angle 6,.

The parameter sensitivity signal can be obtained in various ways e.g. by passing Y,, or Y, through a suitable filter (which would correspond to the closed loop transfer function of the guidance loop). Since the models are not exact replicas ofthe part of the system which they represent, the parameter sensltivity signals will only be approximations to the rate of change and a considerable saving in cost can be obtained by multiplying the adaptive error signals 1, and e, by the algebraic signs of their respective parameter sensitivity values.

in the system shown in FIG. 2, the parameter sensitivity signal for the pitch error is obtained by passing Y through a model 40 of the yaw guidance loop (this model may be identical with the other models or may be a different approximation to the guidance loop) and s, is switched positively or negative ly in the switching unit 41 according to the sign of the parameter sensitivity signal.

The parameter sensitivity signal for the yaw guidance loop is obtained in much the same way and is shown in FIG. 2, the only difference being that the parameter sensitivity signal now needs a sign reversal in an inverter 42 before entering a model 43, the output of which controls a switch 44. An approximate nonmathematical explanation of the action of the parameter sensitivity signals is as follows.

The basic effect of a roll resolver positional error is to cause cross-coupling of the guidance loop i.e. a signal proportional to Y sin (c -6, appears in the pitch loop, and a signal proportional to X sin (B -8, appears in the yaw loop. Thus X s will have two components, one proportional to X and the other proportional to Y(- The component proportional to X,- is equal to X (assuming a perfect match between model and guidance loop) and so the adaptive error s, will be proportional to Y and the roll error (B -(3 )v It might be thought that the roll error alone could be used to change the adaptive resolver angle 9,, to make the error zero but unfortunately Y could change polarity quite indepen' dently of the roll position and the resolver motor would not know whether to turn clockwise or anticlockwise. In order to remove this ambiguity the adaptive error e, is multiplied by a function of Y, i.e. by a signal derived from the opposite loop. The output from model 40 is a function of Y and it can be shown that it is exact phase with the Y component in e, and so the product of these two signals is a function of the square of Y and is thus independent of its polarity. It can now be seen that if e, is multiplied by the sign of the output from 40 the product is still independent of the polarity of Y and the electronic switch 41 is used to achieve this because electronic switches are cheaper and more reliable than multipliers. The outputs of the two switches 41 and 44 are added in a summing amplifier 45 in order to average over the two loops, and so the output from summing amplifier 45 is a signal whose amplitude depends on X and Y but whose sign depends only on the sign of the roll error, Thus, the adaptive resolver will always turn in the right direction but its speed of rotation will depend on the amplitudes of X and Y The signal from amplifier 45 is multiplied by a gain factor G in a circuit 46, passed through a shaping network 47 and then fed into a rate servo 48 which adjusts the resolver until 9,, is equal to 9,, (at which time the input to the rate servo will be zero). Alternatively, the rate servo could be replaced by an integrator and a position servo. Also, if the missile has a large roll rate or is subject to large roll accelerations, additional integrators (with appropriate stabilizing networks) could be included in the adaptive loop immediately prior to the servo to provide zero velocity error or zero acceleration error systems. Finally, an automatic gain device could be included in the adaptive loop to maintain the adaptive loop gain as high as possible consistent with the stability of the system and thus to keep the response of the adaptive roll resolver fast and independent of the amplitude of the guidance loop error signals.

If there is in the missile another resolver based on a gyro measurement of the roll angle, the method of operation is unchanged. The efi'ect of the command signals on the missile will be modified by this resolver and the resolver at the tracking station will therefore now adapt itself, under the control of the loop actuating signals, until it is equal to the difference between the actual missile roll angle 9,, and the gyro measured roll angle 9 In such a case greater accuracy is achieved than would be possible with the missile resolver and gyro, or alternatively, a cheaper roll gyro can be used in the missile.

If, however, the gyro and roll resolver in the missile are dispensed with altogether. the missile system becomes very much cheaper, lighter and less complex.

In a system intended only for use against stationary or very slow moving targets, the models of the pitch and yaw guidance loops which provide the signals X and Y can be omitted, since these signals would be very small, and could be neglected. in such cases the inputs to the switches 4i and 44 would be the signals X; and Y; from the summing amplifiers l8 and 19. The switches 41 and 44 or equivalent devices would still be required to control the polarity of the correction signal applied to the servo 48 and the models 40 and 43 could be retained to provide this polarity information.

Iclaim:

l. A guidance station for remotely guiding a missile, the guidance station comprising: means for generating first signals corresponding to the missile positions in elevation and azimuth; means for generating second signals representing required missile positions in elevation and azimuth; a pitch guidance system responsive to the difference between said first and second elevation signals and a yaw guidance system responsive to the difference between said first and second azimuth signals; a resolver common to said pitch and yaw guidance systems for resolving said elevation and azimuth dif ferences in accordance with an assumed missile roll angle; adaptive means including a pitch circuit responsive to said second elevation signal and to the difference between said first and second elevation signals and a yaw circuit responsive to said second azimuth signal and to the difference between said first and second azimuth signals, said pitch and yaw circuits including models of said guidance systems, said adaptive means including means actuated by said pitch and yaw circuits for adjusting the angle of resolution of said resolver so as to decrease the difference between said first and second signals.

2. A guidance station for remotely guiding a missile, comprising pitch and yaw guidance systems including means for generating pitch and yaw error signals for the correction of missile deviations in elevation and azimuth from a required course, a resolver for resolving the pitch and yaw error signals in accordance with an assumed missile roll angle; means for generating signals representing required elevation and azimuth of said missile, and models of said pitch and yaw guidance systems responsive to said required elevation and azimuth signals; means responsive to said missile pitch and yaw error signals and to corresponding error signals derived from said models to generate pitch and yaw actuating signals; means for modifying the polarity of said pitch and yaw actuating signals in accordance with the yaw and pitch error signals from said models, respectively; and means for adjusting the angle of resolution of said resolver in accordance with said modified signals.

3. A guidance station for remotely guiding a missile comprising: pitch and yaw guidance systems including means for generating pitch and yaw error signals for the correction of missile deviations in elevation and azimuth from a required course, and a resolver for resolving the pitch and yaw error signals in accordance with an assumed missile roll angle; means for generating signals representing required elevation and azimuth angles of said missiles, and models of the pitch and yaw guidance systems connected to receive said signals representing required pitch and yaw angles; means for deriving from said models pitch and yaw error signals corresponding to said missile pitch and yaw error signals; and means for comparing corresponding error signals in each guidance system and in the corresponding model to provide pitch and yaw difference signals; and means actuated by said pitch and to receive the output of said first yaw guidance model and providing an output representing a pitch sensitivity signal. and a further model of said pitch guidance system connected to receive the output of said first pitch guidance model and providing an output representing a yaw sensitivity signal; said guidance station further comprising means for inverting the sign of one of said parameter sensitivity signals, and means for multiplying said difference signals by the algebraic sign of their parameter sensitivity signals.

Images