Fully Symmetric Polytope Set Membership State Estimation Method for Multi-rate Systems
By adopting a fully symmetric multicellular set-member state estimation method based on a dynamic event triggering mechanism, the problem of state estimation for unknown disturbances and measurement noise in multi-rate systems is solved, achieving efficient state point and range estimation while saving communication resources.
Patent Information
- Application Number
- CN202210111985.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-28
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-01-28
AI Technical Summary
Existing state estimation methods struggle to handle unknown but bounded system disturbances and measurement noise in multi-rate systems, and cannot simultaneously achieve efficient state point estimation and range estimation, especially when communication resources are limited under dynamic event-triggered mechanisms.
A fully symmetric polytopic set-member state estimation method based on a dynamic event-triggered mechanism is adopted. By establishing a discrete time-varying linear multi-rate system model, converting it into a single-rate system, introducing a dynamic event-triggered mechanism, and designing a fully symmetric polytopic set-member state estimator through order reduction techniques and parameter optimization, a minimum estimation set containing the true state of the system is obtained.
It realizes state estimation of multi-rate systems under unknown bounded noise conditions, which not only solves the state estimation problem, but also saves communication resources and provides efficient state point estimation and range estimation.
Smart Images

Figure CN114462241B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of information technology, in particular to the field of state estimation, and specifically relates to a fully symmetric polytope set membership state estimation method for a multi-rate system based on a dynamic event triggering mechanism. Background Art
[0002] In the field of state estimation, commonly used state estimation methods include Kalman state estimation method, H ∞ The Kalman state estimation method requires the accurate statistical characteristics of system disturbances and measurement noise to be obtained in advance, but this is difficult to achieve in engineering practice. Therefore, in many practical applications, the Kalman state estimation method has certain limitations. ∞ The state estimation method is applicable to the case where the system disturbance and measurement noise energy are bounded, and can provide the estimation error bound of the system in the worst case. Although the H-infinity state estimation method does not need to be based on the statistical characteristics of the system disturbance and measurement noise, when the system disturbance and measurement noise are unknown but bounded, H ∞ The state method seems to be somewhat insufficient. In addition, the Kalman state estimation method and H ∞ State estimation methods can only make point estimates of the state, but cannot estimate the range to which the state belongs. The ellipsoid-based set membership state estimation method can be applied to the situation where the system disturbance and measurement noise are unknown but bounded, and it is also unnecessary to obtain the statistical characteristics of the system disturbance and measurement noise. However, the ellipsoid-based set membership state estimation method cannot achieve a good trade-off between computational accuracy and computational complexity. In many practical applications, it is difficult to sample system devices and sensors at a consistent rate, so in the state estimation problem, the multi-rate mechanism should be fully considered. Under the dynamic event triggering mechanism, the measurement output will be transmitted to the estimator end only when it meets certain triggering conditions, so unnecessary data transmission can be avoided and limited communication resources can be saved. However, under the dynamic event triggering mechanism, there is no fully symmetric polytope set membership state estimation method for multi-rate systems. Summary of the invention
[0003] The object of the present invention is to provide a method for estimating the state of a fully symmetric polytope set of a multi-rate system based on a dynamic event triggering mechanism.
[0004] The method of the present invention is specifically as follows:
[0005] Step (1) establishing a discrete time-varying linear multi-rate system model;
[0006] Step (2) applies lifting technology to convert the discrete time-varying linear multi-rate system model into a corresponding discrete time-varying linear single-rate system model;
[0007] Step (3) introduces a dynamic event trigger mechanism for the converted discrete time-varying linear single-rate system to save limited communication resources;
[0008] Step (4) designing a fully symmetric polytope set state estimator based on a dynamic event triggering mechanism to obtain a parameterized fully symmetric polytope estimation set containing the true state of the system;
[0009] Step (5) optimizes the parameters to obtain the smallest set of fully symmetric polytope estimates in terms of the F-norm.
[0010] Furthermore, the discrete time-varying linear multi-rate system model established in step (1) is as follows:
[0011] Among them, s is the time sequence number, t s and k s is the sampling time at two rates; t s The system status at the moment, Indicates n x dimensional Euclidean space, t s The system disturbance at time Indicates n w dimensional Euclidean space; k s The measurement output at the moment, Indicates n y dimensional Euclidean space, k s The measurement noise at the moment, Indicates n v dimensional Euclidean space; and is a known time-varying real matrix.
[0012] The sampling period of the system is T′=t s+1 -t s , and t 0 =0; the sampling period of measurement K′=k s+1 -k s , and k 0 =0; cycle ratio is a positive integer greater than or equal to 2.
[0013] Set α system initial states System disturbance and measurement noise Contained in the following holosymmetric polytopes:
[0014] in, and are known vectors and matrices, and They are respectively of order n w and n v The identity matrix of .
[0015] Step (2) is specifically:
[0016] Iteratively apply the discrete time-varying linear multi-rate system model to obtain Among them, the intermediate parameters
[0017] System disturbances for single-rate systems after conversion for The transpose of .
[0018] Intermediate parameters Intermediate parameters
[0019] The time-varying real matrix of the converted single-rate system and for: Intermediate parameters g∈{1,2,…,α};
[0020] System status of single-rate system after conversion for The transpose of .
[0021] The discrete time-varying linear single-rate system model is obtained as follows:
[0022]
[0023] The initial state of the discrete time-varying linear single-rate system model after conversion and system disturbances The following conditions must be met:
[0024] Initial state of discrete time-varying linear single-rate system model after conversion The fully symmetric polytope Among them, the center of the fully symmetric polytope Generator Matrix for Fully Symmetric Polytopes
[0025] Step (3) introduces the following dynamic event triggering mechanism for the converted discrete time-varying linear single-rate system:
[0026] The sequence of triggering moments is 0=ι 0 <ι 1 <…<ι m <…, ιm is the trigger time, m is the trigger time sequence number, m=0,1,…, satisfying: Among them, k s Measurement output at time The most recently transmitted measurement output The difference express Euclidean norm of ; auxiliary dynamic variables satisfy: σ, λ, and θ are given positive scalars, and the initial values of the auxiliary dynamic variables are given If λθ ≥ 1, then for any k s ≥0, yes
[0027] After the dynamic event triggering mechanism is introduced, the input actually received by the set membership estimator is It is expressed as: k s ∈[ι m ,ι m+1 ).
[0028] Step (4) is specifically:
[0029] The estimated set of parameterized fully symmetric polytopes containing the true state of the system obtained by order reduction techniques It has the following structure:
[0030] in, represents a reduction technique that can make the fully symmetric polytope before reduction The order of The generator matrix The number of columns is not greater than the set value r.
[0031] Fully symmetric polytope before reduction
[0032] Select the center of the fully symmetric polytope As a fully symmetric polytope set state estimator based on dynamic event triggering mechanism, it can realize Point estimates of the intermediate parameters
[0033] Generator matrix of fully symmetric polytope before rank reduction Intermediate parameters Generator matrix of the reduced fully symmetric polytope is a parameter matrix of the estimated set of fully symmetric polytopes, i.e., the parameters that need to be optimized.
[0034] Dynamic event triggering mechanism related parameters for:
[0035] in, and is a given positive scalar. Let and So for The upper bound of
[0036] right Perform range estimation and obtain The upper bound of and the lower bound as follows:
[0037] in, The rows and matrices of Intermediate parameters yes The element in the i-th row and q-th column of is a matrix The number of columns.
[0038] Step (5) is specifically:
[0039] Optimized parameters Intermediate parameters Intermediate parameters
[0040] When the fully symmetric polytope estimate set Parameters have In the form of, the set of fully symmetric polytope estimates The F norm of is the smallest, that is, it contains the full symmetric polytope estimation set of the true state of the system The size is the smallest.
[0041] The present invention designs a fully symmetric polytope set membership state estimation method for a multi-rate system based on a dynamic event trigger mechanism. The method can simultaneously handle the sparse data transmission problem caused by the multi-rate mechanism and the dynamic event trigger mechanism, thereby solving the state estimation problem of the multi-rate system and saving limited network communication resources. The method of the present invention assumes that the system disturbance and measurement noise are unknown and bounded, and does not need to determine the statistical characteristics of the system disturbance and measurement noise that are difficult to obtain. Therefore, it is more universal than the traditional estimation method based on the statistical characteristics of the system disturbance and measurement noise; the center of the minimum fully symmetric polytope estimation set containing the real state of the system is used as the point estimate of the state, which is simple and easy to implement; and the value interval of each state in the multi-rate system can be obtained, so as to facilitate the range estimation of the state. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 It is a schematic diagram of the process of the present invention;
[0043] Figure 2 is the system state variable in the embodiment System state variables Estimates System state variables Schematic diagram of the results of the upper and lower bounds of ;
[0044] Figure 3 is the system state variable in the embodiment System state variables Estimates System state variables Schematic diagram of the results of the upper and lower bounds of ;
[0045] Figure 4 In the embodiment, the system state variable The estimated error For system state variables The estimated error Schematic diagram of the results;
[0046] Figure 5 Schematic diagram of the transmission timing of the system measurement output under the dynamic event triggering mechanism in the embodiment;
[0047] Figure 6 It is a comparative schematic diagram of the estimated mean square error when the sensor measurement sampling periods are 2 and 4 respectively in the embodiment. DETAILED DESCRIPTION
[0048] The specific implementation of the present invention is described in detail below in conjunction with the technical scheme and the accompanying drawings.
[0049] To facilitate the subsequent explanation, the following definitions and lemmas are given:
[0050] Holosymmetric polytope definition: a holosymmetric polytope of order m is the hypercube H m =[-1,+1] m Perform the following affine transformation: Where p∈R n is called the center of the fully symmetric polytope X, n is the dimension; J∈R n×m is called the generator matrix of the fully symmetric polytope X. For convenience, let X =<p,J> .
[0051] Lemma: Given a fully symmetric polytope and an integer reduction parameter r (n<r<m), if there is a reduction operator of the form:
[0052]
[0053] in, and yes The element in the i-th row and q-th column of
[0054] like Figure 1 As shown in FIG. 1 , a fully symmetric polytope set state estimation method for a multi-rate system based on a dynamic event triggering mechanism is described as follows:
[0055] Step (1) Establish a discrete time-varying linear multi-rate system model.
[0056] Furthermore, the discrete time-varying linear multi-rate system model established in step (1) is as follows:
[0057] Among them, s is the time sequence number, t s and k s is the sampling time at two rates; t s The system status at the moment, Indicates n x dimensional Euclidean space, t s The system disturbance at time Indicates n w dimensional Euclidean space; k s The measurement output at the moment, Indicates n y dimensional Euclidean space, k s The measurement noise at the moment, Indicates n v dimensional Euclidean space; and is a known time-varying real matrix.
[0058] The sampling period of the system is T′=t s+1 -t s , and t 0 =0; the sampling period of measurement K′=k s+1 -k s , and k 0 =0; cycle ratio is a positive integer greater than or equal to 2.
[0059] Set α system initial states System disturbance and measurement noise Contained in the following holosymmetric polytopes:
[0060] in, and are known vectors and matrices, and They are respectively of order n w and n v The identity matrix of .
[0061] Step (2) applies lifting technology to convert the discrete time-varying linear multi-rate system model into the corresponding discrete time-varying linear single-rate system model. Specifically:
[0062] Iteratively apply the discrete time-varying linear multi-rate system model to obtain Among them, the intermediate parameters
[0063] System disturbances for single-rate systems after conversion for The transpose of .
[0064] Intermediate parameters Intermediate parameters
[0065] Time-varying real matrix of the converted single-rate system and for: Intermediate parameters g∈{1,2,…,α};
[0066] System status of single-rate system after conversion for The transpose of .
[0067] The discrete time-varying linear single-rate system model is obtained as follows:
[0068]
[0069] The initial state of the discrete time-varying linear single-rate system model after conversion and system disturbances The following conditions must be met:
[0070] Initial state of discrete time-varying linear single-rate system model after conversion The fully symmetric polytope Among them, the center of the fully symmetric polytope Generator Matrix for Fully Symmetric Polytopes
[0071] Step (3) introduces a dynamic event trigger mechanism for the converted discrete time-varying linear single-rate system to save limited communication resources.
[0072] For the converted discrete time-varying linear single-rate system, the following dynamic event triggering mechanism is introduced:
[0073] The sequence of triggering moments is 0=ι 0 <ι 1 <…<ι m <…, ι m is the trigger time, m is the trigger time sequence number, m=0,1,…, satisfying: Among them, k s Measurement output at time The most recently transmitted measurement output The difference express Euclidean norm of ; auxiliary dynamic variables satisfy: σ, λ, and θ are given positive scalars, and the initial values of the auxiliary dynamic variables are given If λθ ≥ 1, then for any k s ≥0, yes
[0074] After the dynamic event triggering mechanism is introduced, the input actually received by the set membership estimator is It is expressed as: k s ∈[ι m ,ι m+1 ).
[0075] It can be seen that only Satisfy the trigger conditions The sensor will transmit the current measurement information to the estimator when the sensor is in the state of being measured, otherwise the estimator will continue to use the most recently transmitted measurement output. Therefore, limited communication resources can be saved.
[0076] Step (4) designs a fully symmetric polytope set state estimator based on a dynamic event trigger mechanism to obtain a parameterized fully symmetric polytope estimation set containing the true state of the system. Specifically:
[0077] The estimated set of parameterized fully symmetric polytopes containing the true state of the system obtained by order reduction techniques It has the following structure:
[0078] in, represents a reduction technique that can make the fully symmetric polytope before reduction The order of The generator matrix The number of columns is not greater than the set value r.
[0079] Fully symmetric polytope before reduction
[0080] Select the center of the fully symmetric polytope As a fully symmetric polytope set state estimator based on dynamic event triggering mechanism, it can realize Point estimates of the intermediate parameters
[0081] Generator matrix of fully symmetric polytope before rank reduction Intermediate parameters Generator matrix of the reduced fully symmetric polytope is a parameter matrix of the estimated set of fully symmetric polytopes, i.e., the parameters that need to be optimized.
[0082] Dynamic event triggering mechanism related parameters for:
[0083] in, and is a given positive scalar. Let and So for The upper bound of
[0084] right Perform range estimation and obtain The upper bound of and the lower bound as follows:
[0085] in, The rows and matrices of Intermediate parameters yes The element in the i-th row and q-th column of is a matrix The number of columns.
[0086] Step (5) optimizes the parameters to obtain the smallest set of fully symmetric polytope estimates in terms of the F norm. Specifically:
[0087] Optimized parameters Intermediate parameters Intermediate parameters
[0088] When the fully symmetric polytope estimate set Parameters have In the form of, the set of fully symmetric polytope estimates The F norm (i.e. The generator matrix The F norm of the system is the smallest, that is, the set of fully symmetric polytope estimates containing the true state of the system The size is the smallest.
[0089] The method of the present invention is used to perform simulation verification.
[0090] System parameters:
[0091]
[0092]
[0093] System disturbance Measurement noise
[0094] The sampling period of the system is T′=1, and the sampling period of the measurement is K′=2. The given positive scalars are σ=0.3, λ=0.15, θ=20. The given initial values of the auxiliary dynamic variables are The reduction parameter in the lemma is r=20.
[0095] The initial value of the state is set to: Set α system initial states n∈{0,1,…,α-1}. Initial state Included in the fully symmetric polytope Among them I 2 is the second-order identity matrix.
[0096] The simulation results are as follows Figure 2-6 As shown: Figure 2 Given the system state variables System state variables Estimates System state variables The upper and lower bounds of Figure 3 Given the system state variables System state variables Estimates System state variables The upper and lower bounds of Figure 4 The method of the present invention is used to calculate the system state variables. and The error caused by the estimation and Figure 5 The transmission time of the system measurement output under the dynamic event triggering mechanism is given; Figure 6The estimated mean square errors are compared when the measurement sampling periods are 2 and 4. The simulation results show that the estimated error is bounded and within a reasonable range, which can illustrate the effectiveness of the fully symmetric polytope set membership state estimation method for multi-rate systems based on a dynamic event trigger mechanism proposed in the present invention.
[0097] In summary, the present invention provides a fully symmetric polytope set membership state estimation method based on a dynamic event trigger mechanism for a class of discrete time-varying multi-rate systems under the constraint of unknown but bounded noise. This method can simultaneously deal with the sparse data transmission problems caused by the multi-rate mechanism and the dynamic event trigger mechanism, which not only solves the state estimation problem of such multi-rate systems, but also saves limited communication resources. This method can obtain a parameterized fully symmetric polytope estimation set containing the true state; then, through parameter optimization, the smallest fully symmetric polytope estimation set in the sense of the F norm can be obtained, thereby ensuring excellent estimation accuracy. The center of the smallest fully symmetric polytope estimation set is selected as the point estimate of the state, so that the estimation of the state can be achieved.
Claims
1. A fully symmetric polytope set membership state estimation method for multi-rate systems, characterized in that: The method is as follows: Step (1) establishing a discrete time-varying linear multi-rate system model; Step (2) applies lifting technology to convert the discrete time-varying linear multi-rate system model into a corresponding discrete time-varying linear single-rate system model; specifically: Iteratively apply the discrete time-varying linear multi-rate system model to obtain Among them, the intermediate parameters System disturbances for single-rate systems after conversion for The transpose of Intermediate parameters Intermediate parameters p, f∈{1, 2, …, α}; The time-varying real matrix of the converted single-rate system and for: Intermediate parameters, System status of single-rate system after conversion for The transpose of The discrete time-varying linear single-rate system model is obtained as follows: The initial state of the discrete time-varying linear single-rate system model after conversion and system disturbances The following conditions must be met: Initial state of discrete time-varying linear single-rate system model after conversion The fully symmetric polytope Among them, the center of the fully symmetric polytope Generator Matrix for Fully Symmetric Polytopes Step (3) introduces a dynamic event trigger mechanism for the converted discrete time-varying linear single-rate system to save limited communication resources; Step (4) designing a fully symmetric polytope set state estimator based on a dynamic event triggering mechanism to obtain a parameterized fully symmetric polytope estimation set containing the true state of the system; Step (5) optimizes the parameters to obtain the smallest set of fully symmetric polytope estimates in terms of the F-norm.
2. The method for estimating state of a fully symmetric polytope set for a multi-rate system according to claim 1, wherein: The discrete time-varying linear multi-rate system model established in step (1) is as follows: Among them, s is the time sequence number, t s and k s is the sampling time at two rates; t s The system status at the moment, Indicates n x dimensional Euclidean space, t s The system disturbance at time Indicates n w dimensional Euclidean space; k s The measurement output at the moment, Indicates n y dimensional Euclidean space, k s The measurement noise at the moment, Indicates n v dimensional Euclidean space; and is a known time-varying real matrix; The sampling period of the system is T′=t s+1 -t s , and t0 = 0; the measured sampling period K′ = k s+1 -k s , and k0=0; the period ratio is a positive integer greater than or equal to 2; Set α system initial states n∈{0, 1, …, α-1}, system disturbance and measurement noise Contained in the following holosymmetric polytopes: in, and are known vectors and matrices, and They are respectively of order n w and n v The identity matrix of .
3. The method for estimating state of a fully symmetric polytope set for a multi-rate system as claimed in claim 2, characterized in that: Step (3) introduces the following dynamic event triggering mechanism for the converted discrete time-varying linear single-rate system: The sequence of triggering moments is 0=ι0<ι1<…<ι m <…, ι m is the triggering moment, m is the triggering moment sequence number, m=0, 1,…, satisfying: Among them, k s Measurement output at time The most recently transmitted measurement output The difference express Euclidean norm of ; auxiliary dynamic variables satisfy: σ, λ, and θ are given positive scalars, and the initial values of the auxiliary dynamic variables are given If λθ ≥ 1, then for any k s ≥0, yes After the dynamic event triggering mechanism is introduced, the input actually received by the set membership estimator is It is expressed as: k s ∈[ι m ,ι m+1 ).
4. The method for estimating state of a fully symmetric polytope set for a multi-rate system according to claim 3, wherein: Step (4) is specifically: The estimated set of parameterized fully symmetric polytopes containing the true state of the system obtained by order reduction techniques It has the following structure: in, represents a reduction technique that can make the fully symmetric polytope before reduction The order of The generator matrix The number of columns is not greater than the set value r; Fully symmetric polytope before reduction Select the center of the fully symmetric polytope As a fully symmetric polytope set state estimator based on dynamic event triggering mechanism, it can realize Point estimates of the intermediate parameters Generator matrix of fully symmetric polytope before rank reduction Intermediate parameters Generator matrix of the reduced fully symmetric polytope is a parameter matrix of the fully symmetric polytope estimation set, i.e., the parameters that need to be optimized; Dynamic event triggering mechanism related parameters for: in, and is a given positive scalar; let and So for The upper bound of right Perform range estimation and obtain The upper bound of and the lower bound as follows: in, The rows and matrices of Intermediate parameters yes The element in the i-th row and q-th column of is a matrix The number of columns.
5. The method for estimating state of a fully symmetric polytope set for a multi-rate system according to claim 4, wherein: Step (5) is specifically: Optimized parameters Intermediate parameters Intermediate parameters When the fully symmetric polytope estimate set Parameters have In the form of, the set of fully symmetric polytope estimates The F norm of is the smallest, that is, it contains the full symmetric polytope estimation set of the true state of the system The size is the smallest.
Citation Information
Patent Citations
State estimation method for event-triggered transmission complex network based on set membership estimation
CN112260867A
System state estimation method for multi-sensor information fusion under protocol scheduling
CN113271222A
Optimized pulsatile-flow ventricular-assist device and total artificial heart
US20030032853A1
