CN102297771A - Progressive health monitoring method for cable system based on angle monitoring during supporting-base angular displacement - Google Patents

Abstract

The progressive health monitoring method of the cable system based on angle monitoring when the angular displacement of the support is based on angle monitoring, by monitoring the angular coordinates of the structural support to determine whether it is necessary to update the mechanical calculation benchmark model of the structure again, taking into account the current value of the monitored quantity The vector is similar to the linear relationship between the initial value vector of the monitored quantity, the unit damage monitored quantity change matrix and the current nominal damage vector. method, divide the large interval into continuous small intervals, the above-mentioned linear relationship is accurate enough in each small interval, and the current cable damage vector can be calculated by using a suitable algorithm such as a multi-objective optimization algorithm in each small interval According to the non-inferior solution of , the location of the damaged cable and its damage degree can be determined more accurately.

Description

Progressive health monitoring method of cable system based on angle monitoring during bearing angular displacement

technical field

Cable-stayed bridges, suspension bridges, truss structures and other structures have one thing in common, that is, they have many parts that bear tensile loads, such as cable-stayed cables, main cables, slings, tie rods, etc. , cables, or rods that only bear tensile loads are supporting components, and this type of structure is expressed as a "cable structure" in the present invention for convenience. When there is an angular displacement of the support (for example, the rotation of the support around the coordinate axes X, Y, and Z is actually the angular displacement of the support around the coordinate axes X, Y, and Z), the invention discloses a progressive method . The support system for identifying cable structures based on angle monitoring (referring to all load-bearing cables and all supporting rods that only bear tensile loads. For convenience, this patent refers to all support components of this type of structure as " cable system", but in fact the cable system not only refers to the supporting cable, but also includes the damaged cable in the member that only bears the tensile load) (for the truss structure, it refers to the damaged member that only bears the tensile load) The method belongs to the field of engineering structure health monitoring.

Background technique

The cable system is usually a key component of the cable structure, and its failure often leads to the failure of the entire structure. Based on the structural health monitoring technology to identify the damaged cable in the cable system of the cable structure (as mentioned earlier, it also refers to the cable that only bears tension. loaded member) is a promising approach. Changes in the health status of the cable system will cause changes in the measurable parameters of the structure, such as changes in the angular coordinates of any imaginary straight line passing through each point of the cable structure (such as any one of the tangent planes at any point on the surface of the structure the change of the angular coordinate of the straight line passing through the point, or the change of the angular coordinate of the normal line at any point on the surface of the structure), in fact, the change of the structural angle contains the health status information of the cable system, that is to say, the structural angle data can be used to judge The health status of the structure can be based on angle monitoring (this invention refers to the monitored angle data as "monitored quantity", and the "monitored quantity" mentioned later refers to the monitored angle data) to identify damaged cables, and the monitored In addition to being affected by the health status of the cable system, the quantity is also affected by the angular displacement of the cable structure support (which often occurs). At present, there is no public and effective health monitoring system and method to solve this problem.

When there is angular displacement of the support, in order to have a reliable monitoring and judgment on the health state of the cable system of the cable structure, there must be a reasonable and effective establishment of the change of the angle data of the cable structure and the health of all cables in the cable system.

The method of the relationship between conditions, the health monitoring system based on this method can give a more credible health assessment of the cable system.

Contents of the invention

Technical problem: The purpose of this invention is to provide a method based on angle monitoring that can reasonably and effectively monitor the progress of the cable system in the cable structure for the health monitoring problem of the cable system in the cable structure when the support of the cable structure has angular displacement. method of health monitoring.

Technical solution: the present invention consists of two parts. They are: 1. The method of establishing the knowledge base and parameters required by the cable system health monitoring system, and the progressive health of the cable system based on the knowledge base (including parameters) and the angle of the cable structure measured and the angular displacement of the support of the cable structure. State assessment method; Second, the software and hardware parts of the health monitoring system.

The first part of the present invention: the method for establishing the knowledge base and parameters required by the cable system health monitoring system, and the cable system progressive formula based on the knowledge base (including parameters) and the angle of the actually measured cable structure and the angular displacement of the actual measured cable structure support health status assessment methods. The following steps can be followed in a cyclical and progressive manner to obtain a more accurate evaluation of the health status of the cable system.

Step 1: At the beginning of each cycle, it is first necessary to establish or have established the initial damage vector d _o ⁱ of the cable system at the beginning of this cycle ( i =1, 2, 3,…), and establish the initial mechanical calculation basis of the cable structure Model A _o (such as the finite element benchmark model, A _o is constant in the present invention), establish the current mechanical calculation benchmark model A ^ti _o of the cable structure (such as the finite element benchmark model, in each cycle A ^ti _o is Constantly updated), establish the mechanical calculation benchmark model A ⁱ of the cable structure (such as the finite element benchmark model, i =1, 2, 3,…). Except where the letter i clearly represents the step number, in the present invention, the letter i only represents the number of cycles, that is, the i-th cycle.

Assuming that there are N cables in the cable system, the initial damage vector of the cable system required at the beginning of the i-th cycle is recorded as d _o ⁱ (as shown in formula (1)), and d _o ⁱ represents the cable structure at the beginning of the cycle ( The health state of the cable system is represented by the mechanical calculation benchmark model A ⁱ ).

                 (1)

(1)

In formula (1), d ⁱ _oj ( i =1, 2, 3,… ; j =1, 2, 3, ……, N ) represents ^the The initial damage value of the jth cable of the system. When d ⁱ _oj is 0, it means that the jth cable has no damage; when it is 100%, it means that the cable has completely lost its bearing capacity; The cable loses a corresponding proportion of its bearing capacity.

When the initial damage vector of the cable system is ^established _at the beginning of the first cycle (recorded as d ¹ _o according to Eq. . If there is no non-destructive testing data of the cable and other data that can express the healthy state of the cable, or when the initial state of the structure can be considered as a non-damaged state, the value of each element of the vector d ¹ _o is 0.

The initial damage vector d ⁱ _o of the cable system required at the beginning of the i-th cycle ( i =2, 3, 4, 5, 6…) is in the previous cycle (that is, the i-1th cycle, i =2, 3, 4 , 5, 6…) calculated before the end of the cycle, the specific method will be described later.

The mechanical calculation benchmark model that needs to be established at the beginning of the i-th cycle or the established mechanical calculation benchmark model is denoted as A ⁱ .

According to the actual measurement data of the cable structure when the cable structure is completed (including non-destructive testing data of the cable and other data that can express the health status of the cable, cable structure shape data, cable force data, pull rod tension data, cable structure support coordinate data, cable Structural support angle coordinate data, cable structure modal data and other measured data, for cable-stayed bridges and suspension bridges are bridge type data, cable force data, bridge modal data, design drawings, as-built drawings, using Mechanical methods (such as finite element method) to establish A _o ; if there is no actual measurement data of the cable structure when it is completed, then the actual measurement of the structure is carried out before the health monitoring system is established to obtain the actual measurement data of the cable structure (including the shape data of the cable structure , cable force data, pull rod tension data, cable structure support coordinate data, cable structure support angular coordinate data, cable structure modal data and other measured data, for cable-stayed bridges and suspension bridges, it is the bridge type data, cable force Data, bridge modal data, cable non-destructive testing data and other data that can express the health status of the cable), according to this data and the design drawing and as-built drawing of the cable structure, use mechanical methods (such as finite element method) to establish A _o . Regardless of the method used to obtain A _o , the calculated data of the cable structure based on A _o (for cable-stayed bridges and suspension bridges, the bridge type data, cable force data, and bridge modal data) must be very close to the measured data. Data, the error is generally not greater than 5%. In this way, the strain calculation data, cable force calculation data, cable structure shape calculation data, displacement calculation data, cable structure angle data, etc. under the simulated situation obtained by A _o calculation can be reliably close to the measured data when the simulated situation actually occurs . The angular coordinate data of the cable structure support corresponding to A _o compose the initial cable structure support angular coordinate vector U _o . A _o and U _o are constant and only established at the beginning of the first cycle.

The basic model for mechanical calculation of the cable structure established at the beginning of the first cycle is denoted as A ¹ , and A ¹ is equal to A _o .

The mechanical calculation benchmark model A ⁱ required at the beginning of the i-th ( i =2, 3, 4, 5, 6…) cycle is the previous (i-1th, i =2, 3, 4, 5 , 6…) calculated before the end of the cycle, the specific method will be described later.

After the mechanical calculation benchmark model A ¹ and the initial damage vector d ¹ _o of the cable system are available, the damage of each cable in the model A ¹ is expressed by the vector d ¹ _o . On the basis of A ¹ , the damage of all cables is changed to 0, and the mechanical model A ¹ is updated to a mechanical model in which the damage of all cables is 0 (denoted as A ⁰ ), and the mechanical model A ⁰ is actually intact The mechanical model corresponding to the cable structure. The model A ⁰ may be called the damage-free model A ⁰ of the cable structure.

"All monitored angle data of the structure" is described by K specified points on the structure, L specified straight lines passing through each specified point, and H angular coordinate components of each specified straight line. The change of the structure angle is all The change in all specified angular coordinate components of all specified lines for the specified point. Each time there are M (M=K×L×H) angle coordinate component measured or calculated values to represent the angle information of the structure. K and M shall not be less than the number N of supporting cables. For the sake of convenience, the "monitored angle data of the structure" is simply referred to as "monitored quantity" in the present invention.

The present invention uses "initial numerical vector C ⁱ _o of the monitored quantity" ( i =1, 2, 3,...) to represent all The initial value of the specified monitored quantity (see formula (2)), the full name of C ⁱ _o is "the initial value vector of the monitored quantity in the i-th cycle".

              (2)

(2)

In formula (2), C ⁱ _ok ( i =1, 2, 3,…; k =1, 2 , 3, … ., M; M≥N; ) is a monitored quantity. The vector C ⁱ _o is formed by arranging the M monitored quantities defined above according to a certain order, and there is no special requirement for the order of arrangement, it is only required that all related vectors also arrange the data in this order.

the

At the beginning of the first cycle, "the initial value vector C ¹ _o of the monitored quantity in the first cycle" (see formula (2)) is composed of measured data, since the initial value of the monitored quantity calculated according to the model A ¹ is reliably Close to the corresponding actual measured value, in the following description, the same symbol will be used to represent the calculated value component vector and the actual measured value component vector.

The "initial numerical vector C ⁱ _o of the monitored quantity of the i-th cycle" required at the beginning of the i-th cycle ( i =2, 3, 4, 5, 6...) is in the previous (that is, the i-1th , i =2, 3, 4, 5, 6…) calculated before the end of the cycle, the specific method will be described later.

Step 2: During the service process of the cable structure, in each cycle, the current data of the support angle coordinates of the cable structure are continuously measured (all data form the current measured support angle coordinate vector U ^ti of the cable structure, and the definition method of the vector U ^ti Same as vector U _o ). For the sake of convenience, for the i-th cycle, the current data of the angular coordinates of the cable structure support when the current mechanical calculation benchmark model was updated last time is recorded as the current angular coordinate vector U ^ti _o of the cable structure support. The method of establishing and updating A ^ti _o is: at the beginning of each cycle, the current mechanical calculation benchmark model A ^ti _o of the cable structure is equal to A ⁱ ( i = 1, 2, 3, 4, 5, 6...). During the service process of the cable structure, the support angle coordinate data of the cable structure is continuously measured to obtain the current measured support angle coordinate vector U ^ti of the cable structure. If U ^ti is equal to U ^ti _o , there is no need to update A ^ti _o ; if U ^ti is not equal to U ^ti _o , then A ^ti _o needs to be updated. At this time, the difference between U ^ti and U _o is the support angular displacement of the cable structure support with respect to the initial position (corresponding to A _o ) (using the support angular displacement Vector V represents the angular displacement of the support). The method to update A ^ti _o is: on the basis of A _o , let the health status of the cable be the initial damage vector d ⁱ _o of the cable system, and then further impose the current support angular displacement constraint on the cable structure support in A _o , the current support The value of the seat angular displacement constraint is taken from the value of the corresponding element in the current support angular displacement vector V. After applying the current support angular displacement constraint to the cable structure support in A _o , the final result is the updated current mechanical calculation basis For the model A ^ti _o , when A ^ti _o is updated, all element values of U ^ti _o are also replaced by all element values of U ^ti , that is, U ^ti _o is updated, so that U ^ti _o corresponding to A ^ti _o is obtained correctly.

the

Step 3: Each cycle needs to establish the "change matrix of monitored quantity of unit damage " and "nominal unit damage vector", and the "change matrix of monitored quantity of unit damage" established in the ith cycle is recorded as ΔC ⁱ twenty three,…). The "nominal unit damage vector" established in the i-th cycle is denoted as D ⁱ _u . In each cycle, ΔC ⁱ and D ⁱ _u are constantly updated, that is, while updating the current mechanical calculation benchmark model A ^ti _o , the cable structure unit damage monitored variable matrix ΔC ⁱ and the nominal unit damage vector are updated as D ⁱ _u .

The process of establishing and updating the cable structure unit damage monitored quantity change matrix ΔC ⁱ and the nominal unit damage vector” denoted as D ⁱ _u is as follows:

Several calculations are performed on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, and the number of calculations is numerically equal to the number of all cables. Each calculation assumes that there is only one cable in the cable system, and the unit damage (for example, 5%, 10%, 20% or 30% etc.) is added on the basis of the original damage (the original damage can be 0 or not). Damage is unit damage). For the convenience of calculation, when setting the unit damage in each cycle, the structural health state at the beginning of the cycle can be regarded as completely healthy, and the unit damage can be set on this basis (in the subsequent steps, the calculated , the damage value of the cable --- called the nominal damage d ⁱ _c ( i =1, 2, 3,...), all are relative to the beginning of this cycle, when the healthy state of the cable is considered to be completely healthy Therefore, the calculated nominal damage must be converted into real damage according to the formula given later.). The damaged cable in each calculation of the same cycle is different from the damaged cable in other calculations, and the unit damage value of each assumed damaged cable can be different from the unit damage value of other cables, using "nominal unit damage The vector D ⁱ _u ” (shown in formula (3)) records the assumed unit damage of all cables in each cycle, which is recorded as D ¹ _u in the first cycle, and each calculation uses mechanical methods (such as finite element method ) to calculate the current calculated values of the M monitored quantities that have been specified in front of the cable structure, and the current calculated values of the M monitored quantities obtained from each calculation form a "calculated current value vector of the monitored quantities" (when it is assumed that the first When there is a unit damage to the j root cable, formula (4) can be used to express the calculated current value vector C ¹ _tj of all the specified M monitored quantities; the calculated current value vector of the monitored quantity obtained by each calculation minus the monitored quantity The initial value vector C ¹ _o , the resulting vector is the "value change vector of the monitored quantity" under this condition (marked by the position or number of the cable with unit damage) (when the jth cable has unit damage, Use δC ¹ _j to represent the numerical change vector of the monitored quantity, the definition of δC ¹ _j is shown in formula (5), formula (6) and formula (7), formula (5) is formula (4) minus formula (2) Divided by the jth element D _uj of the vector D ¹ _u ), each element of the value change vector δC ¹ _j of the monitored quantity represents the cable that is assumed to have unit damage during calculation (for example, the jth cable) The unit damage (such as D _uj ), and the change rate of the value change of the monitored quantity corresponding to the element relative to the assumed unit damage D _uj ; there are N "monitored quantity values" Change vector", each value change vector of the monitored quantity has M (generally, M≥N ) elements, and these N "value change vectors of the monitored quantity" in turn form a "unit" with M×N elements The change matrix of damage monitored quantity ΔC ¹ ” (M rows and N columns ) , each vector δC ¹ _j ( j =1, 2, 3, ……., N ) is a column of matrix ΔC ¹ , and the definition of ΔC ¹ is as follows: (8) shown.

          (3)

(3)

The element D ⁱ _uj ( i =1, 2, 3,…; j =1, 2, 3, ……., N ) of the nominal unit damage vector D ⁱ _u in formula ( 3 ) represents the assumed The unit damage value of the jth cable, the value of each element in the vector D ⁱ _u can be the same or different.

(4)

Element C ⁱ _tjk in formula (4) ( i =1, 2, 3, … ; j = 1, 2, 3, …., N; k = 1, 2, 3, …., M; M≥ N ) represents the calculated current value of the k- th specified monitored quantity corresponding to the numbering rule when the j -th cable has unit damage in the i- th cycle.

(5)

The superscript i ( i =1, 2, 3, ...) of each quantity in formula (5) indicates the i -th cycle, and the subscript j ( j =1, 2, 3, ...., N ) indicates the j -th root Cable has unit damage, where D ⁱ _uj is the jth element in the vector D ⁱ _u . The definition of vector δC ⁱ _j is shown in formula (6). The kth ( k =1, 2, 3, .... , M; M≥N ) element δC ⁱ _jk of δC ⁱ _j represents the i- th cycle , when the matrix ΔC ⁱ is established, assuming that the j -th cable has unit damage, the calculated change rate of the k -th monitored quantity relative to the assumed unit damage D ⁱ _uj is defined as shown in formula (7).

      (6)

(6)

                                       (7)

(7)

The definitions of the quantities in formula (7) have been described above.

(8)

In the formula (8), the vector δC ⁱ _j ( i =1, 2, 3, …….,, j =1, 2, 3, ……., N ) means that in the i -th cycle, since the j -th cable has The relative value change of all monitored quantities caused by unit damage D ⁱ _uj . The numbering rule of the column (subscript j ) of the matrix ΔC ⁱ is the same as the numbering rule of the subscript j of the elements of the previous vector d ⁱ _o .

During the service process of the cable structure, in each cycle, the current data of the support angle coordinates of the cable structure are continuously measured. Once U ^ti is not equal to U ^ti _o , it is necessary to go back to the second step to update A ^ti _o , After updating A ^ti _o , enter this step to update ΔC ⁱ . In fact, ΔC ⁱ is constantly updated in each cycle, that is, after updating the current mechanical calculation benchmark model A ^ti _o , the change matrix ΔC ⁱ of the monitored damage of the cable structure unit is updated.

the

Step 4: Identify the current health status of the cable system. The specific process is as follows.

In the ith cycle ( i =1, 2, 3, ...), the cable system "the current (calculated or measured) value vector C ⁱ of the monitored quantity" is the same as the "initial value vector C ⁱ _o of the monitored quantity", " The approximate linear relationship between the unit damage monitored quantity change matrix ΔC ⁱ ” and the “current nominal damage vector d ⁱ _c ” is shown in formula (9) or formula (10).

                                  (9)

(9)

                                  (10)

(10)

The definition of the current (calculated or measured) numerical vector C ⁱ of the monitored quantity in formulas (9) and (10) is similar to the definition of the initial numerical vector C ⁱ _o of the monitored quantity, see formula (11); The definition of the nominal damage vector d ⁱ _c is shown in formula (12).

(11)

The element C ⁱ _k ( i =1, 2, 3, …….; k =1, 2, 3, …., M; M≥N ) in the formula (11) is the cable structure of the i -th cycle, The current value of the monitored quantity with the number k corresponding to the numbering rule.

              (12)

(12)

In formula (12), d ⁱ _cj ( i =1, 2, 3,…….; j =1, 2, 3,……., N ) is the current nominal Damage value, the numbering rule of the subscript j of the element of the vector d ⁱ _c is the same as the numbering rule of the column of the matrix ΔC ⁱ .

When the actual damage of the cable is not too large, since the cable structure material is still in the linear elastic stage, the deformation of the cable structure is also small, and the error of such a linear relationship represented by formula (9) or formula (10) is relatively small compared with the actual situation. Small, the error can be defined by the error vector e ⁱ (Equation (13)), which expresses the error of the linear relationship shown in Equation (9) or Equation (10).

(13)

In formula (13), abs() is an absolute value function, and the absolute value is taken for each element of the vector obtained in the brackets.

Since there is a certain error in the linear relationship represented by formula (9) or formula (10), it cannot be solved directly based on formula (9) or formula (10) and "the current (actually measured) value vector C ⁱ of the monitored quantity". Get the current nominal damage vector d ⁱ _c of the cable. If this is done, the elements in the obtained damage vector d ⁱ _c may even have relatively large negative values, that is, negative damage, which is obviously unreasonable. Therefore, it is a reasonable solution to obtain an acceptable solution of the cable damage vector d ⁱ _c (that is, with a reasonable error, but the position of the damaged cable and its damage degree can be determined more accurately from the cable system), and the formula ( 14) to express this method.

                              (14)

(14)

In Equation (14), abs() is an absolute value function, and the vector g ⁱ describes the deviation from the ideal linear relationship (Equation (9) or Equation (10))

The reasonable deviation of is defined by formula (15).

                 (15)

(15)

g ⁱ _k ( i =1, 2, 3, ….; k =1, 2, 3, …., M ) in formula (15 ) describes the deviation from formula (9) or formula ( 10) The maximum permissible deviation from the ideal linear relationship shown. The vector g ⁱ can be selected according to the error vector e ⁱ defined by formula (13).

When the initial numerical vector C ⁱ _o of the monitored quantity (obtained from actual measurement or calculation), the change matrix of the monitored quantity of cable structure unit damage ΔC ⁱ (obtained from calculation) and the current numerical vector C ⁱ of the monitored quantity (obtained from actual measurement) are known , you can use a suitable algorithm (such as a multi-objective optimization algorithm) to solve equation (14) to obtain an acceptable solution for the current nominal damage vector d ⁱ _c of the cable system, and the current actual damage vector d ⁱ of the cable system (see equation (16) for definition ) elements can be calculated according to formula (17), that is, the current actual damage vector d ⁱ of the cable is obtained, so that the position and damage degree of the damaged cable can be determined by d ⁱ , that is, the health monitoring of the cable system is realized.

                 (16)

(16)

In formula (16), d ⁱ _j ( i =1, 2, 3,… ; j =1, 2, 3, ……., N ) represents the actual damage value of the jth cable in the i -th cycle, and its definition See formula (17), when d ⁱ _j is 0, it means that the jth cable has no damage; when it is 100%, it means that the cable completely loses its bearing capacity; when it is between 0 and 100%, it means that the jth cable loses the corresponding proportion Bearing capacity, the numbering rule of the elements of the vector d ⁱ is the same as that of the elements of the vector d ⁱ _o in formula (1).

                                   (17)

(17)

In formula (17), d ⁱ _oj ( i =1, 2, 3, 4, … ; j =1, 2, 3, ……., N ) is the jth element of vector d ⁱ _o , and d ⁱ _cj is The jth element of the vector d ⁱ _c .

Step 5: Determine whether to end this ( i -th) cycle, if yes, complete the finishing work before the end of this cycle, for the next time (that is, i + 1 time, i = 1, 2, 3, 4 , …) loop to prepare the benchmark model and necessary vectors for the mechanics calculations. The specific process is as follows.

After obtaining the current nominal damage vector d ⁱ _c in this ( i- th) cycle, first, establish the identification vector F ⁱ according to formula (18), and formula (19) gives the jth element of the identification vector F ⁱ definition; if the elements of the identification vector F ⁱ are all 0, then continue to monitor and calculate the health of the cable system in this cycle; if the elements of the identification vector F ⁱ are not all 0, after completing the subsequent steps, enter the next cycle. The so-called subsequent steps are: first, calculate the initial damage vector d ^{i + 1} _o required for the next ( i + 1th, i = 1, 2, 3, 4, ...) cycle according to formula (20) Each element d ^{i + 1} _oj ; secondly, on the basis of the mechanical calculation benchmark model A ⁱ ( i =1, 2, 3, 4, …) or the damage-free model A ⁰ of the cable structure, let the health status of the cable After the state is d ^{i + 1} _o, the mechanical calculation benchmark model A ⁱ⁺¹ required for the next (i+1th, i = 1, 2, 3, 4, ...) cycle is updated; finally, through the mechanical calculation The calculation of the benchmark model A ⁱ⁺¹ obtains the initial value of the monitored quantity, which constitutes the "monitored quantity's Initial numerical vector C ⁱ⁺¹ _o ” ( i =1, 2, 3, 4, …).

                 (18)

(18)

The superscript i of the identification vector F ⁱ in formula (18) represents the i- th cycle, and the subscript j of its element F ⁱ _j ( j = 1, 2, 3, ..., N) represents the damage characteristics of the j -th cable, Only two quantities of 0 and 1 can be taken, and the specific value rules are shown in formula (19).

                                   (19)

(19)

In formula (19), the element F ⁱ _j is the jth element of the identification vector F ⁱ , D ⁱ _uj is the jth element of the nominal unit damage vector D ⁱ _u (see formula (3)), and d ⁱ _cj is the cable system The jth element of the current nominal damage vector d ⁱ _c (see formula (12)), they all represent the relevant information of the jth root cable.

                               (20)

(20)

In formula (20), D ⁱ _uj is the jth element of the nominal unit damage vector D ⁱ _u (see formula (3)), and d ⁱ _cj is the jth element of the current nominal damage vector d ⁱ _c of the cable system (see formula (12)).

the

The second part of the present invention: the software and hardware parts of the health monitoring system.

The hardware part includes the monitoring system (including the monitored quantity monitoring system, the cable structure support angular coordinate monitoring system), signal collector and computer, etc. Real-time or quasi-real-time monitoring of each monitored quantity is required, and real-time or quasi-real-time monitoring of the angular coordinates of each cable structure support is required.

The software part should be able to complete the process set in the first part of the present invention, that is, to complete the monitoring, recording, control, storage, calculation, notification, alarm and other functions required in the present invention that can be realized by computer.

the

The inventive method specifically comprises:

a. Assuming that there are N cables in total, first determine the numbering rules of the cables, and number all the cables in the cable structure according to this rule, and the numbers will be used to generate vectors and matrices in subsequent steps;

b. Determine the specified measured point, and number all the specified points; determine the measured straight line of each measured point, and number all the specified measured straight lines; determine the measured angular coordinate component of each measured straight line, and give The numbers of all measured angular coordinate components. The above numbers will be used in subsequent steps to generate vectors and matrices. The "all monitored angular data of the structure" consists of all the above-mentioned measured angular coordinate components. For the sake of convenience, the "monitored angle data of the structure" is simply referred to as "monitored quantity" in the present invention. The number of measuring points shall not be less than the number of cables; the sum of the numbers of all measured angle coordinate components shall not be less than the number of cables;

c. Establish the initial damage vector d ¹ _o of the cable system by using the nondestructive testing data of the cable and other data that can express the healthy state of the cable; if there is no nondestructive testing data of the cable and other data that can express the healthy state of the cable, the structure When the initial state is a non-damaged state, the value of each element of the vector d ¹ _o is 0;

d. While establishing the initial damage vector d ¹ _o of the cable system, directly measure and calculate all the specified monitored quantities of the cable structure to form the "initial value vector C ¹ _o of the monitored quantity";

e. While establishing the initial damage vector d ¹ _o of the cable system and the initial numerical vector C ¹ _o of the monitored quantity, the initial cable force data of all cables of the cable structure are obtained through actual measurement, and the initial geometric data of the cable structure are obtained through actual measurement;

f. Establish the initial mechanical calculation benchmark model A _o of the cable structure, establish the initial cable structure support angle coordinate vector U _o , and establish the mechanical calculation benchmark model A ¹ of the cable structure required at the beginning of the first cycle; The actual measured data of the cable structure, the measured data includes cable structure shape data, cable force data, pull rod tension data, cable structure support coordinate data, cable structure support angular coordinate data, cable structure modal data and other measured data, the cable structure The non-destructive testing data and other data that can express the health state of the cable, according to the design drawing and the as-built drawing, use the mechanical method to establish the initial mechanical calculation benchmark model A _o of the cable structure; if there is no actual measurement data of the cable structure when it is completed, then Just before the establishment of the health monitoring system, the actual measurement of the cable structure is carried out, and the actual measurement data of the cable structure is also obtained. According to this data and the design drawing and completion drawing of the cable structure, the initial mechanical calculation benchmark model A of the cable structure is also established by using the mechanical method _. ; Regardless of the method used to obtain A _o , the calculated data of the cable structure based on A _o must be very close to the measured data, and the difference between them must not be greater than 5%; the angle coordinate data of the cable structure corresponding to A _o constitute the initial cable Structural support angular coordinate vector U _o ; A _o and U _o are invariable, and are only established at the beginning of the first cycle; the mechanical calculation benchmark model of the cable structure established at the beginning of the i-th cycle is denoted as A ⁱ , where i represents the number of cycles; in the application of the present invention, the letter i only represents the number of cycles, i.e. the i-th cycle, except where the letter i clearly represents the step number in the application of the present invention; therefore the mechanical calculation of the cable structure established when the first cycle begins The reference model is denoted as A ¹ , and A ¹ is equal to A _o in the present invention; for the convenience of description, it is named "Cable Structure Current Mechanics Calculation Reference Model A ^ti _o ", and A ^ti _o will be continuously updated in each cycle according to the need, and every At the beginning of a cycle, A ^ti _o is equal to A ⁱ ; also for the convenience of description, it is named "cable structure measured support angle coordinate vector U ^ti ", in each cycle, the current data of the cable structure support angle coordinates are continuously measured, all The current data of the support angle coordinates of the cable structure form the current measured support angle coordinate vector U ^ti of the cable structure. The elements of the vector U ^ti and the elements at the same position as the vector U _o represent the angular coordinates of the same support in the same direction; for the convenience of description, For the ith cycle, the current data of the cable structure support angle coordinates when A ^ti _o was updated last time is recorded as the current cable structure support angle coordinate vector U ^ti _o ; at the beginning of the first cycle, A ^t1 _o is equal to A ¹ , U ^t1 _o is equal to U _o ;

g. At the beginning of each cycle, let A ^ti _o be equal to A ⁱ ; the current data of the cable structure support angular coordinates are obtained from the actual measurement, and all the current data of the cable structure support angular coordinates form the current cable structure measured support angular coordinate vector U ^ti , according to The measured support angle coordinate vector U ^ti of the current cable structure, and update the current mechanical calculation benchmark model A ^ti _o of the cable structure and the current support angle coordinate vector U ^ti _o of the cable structure when necessary;

h. Carry out several mechanical calculations on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, and obtain the cable structure unit damage monitored quantity change matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u through calculation;

i. The current measured values of all the specified monitored quantities of the cable structure are obtained through actual measurement, and form the "current value vector C ⁱ of the monitored quantities". When numbering the elements of all vectors that appear in this step and before this step, the same numbering rule should be used, so as to ensure that the elements with the same number in each vector that appeared in this step and before this step represent the corresponding value of the same monitored quantity. related information defined in the vector to which the element belongs;

j. Define the current nominal damage vector d ⁱ _c and the current actual damage vector d ⁱ of the cable system. The number of elements in the damage vector is equal to the number of cables. There is a one-to-one correspondence between the elements of the damage vector and the cables. The value of the elements of the damage vector Represents the damage degree or health status of the corresponding cable;

k. According to the relationship between "the current numerical vector C ⁱ of the monitored quantity" and "the initial numerical vector C ⁱ _o of the monitored quantity", "the change matrix ΔC ⁱ of the monitored quantity with unit damage" and "the current nominal damage vector d ⁱ _c " Existing approximate linear relationship, the approximate linear relationship can be expressed as formula 1, in formula 1, other quantities except d ⁱ _c are known, and the current nominal damage vector d ⁱ _c can be calculated by solving formula 1;

                                式1

Formula 1

l. Using the relationship between the current actual damage vector d i expressed in formula 2 ^and the elements of the initial damage vector d ⁱ _o and the current nominal damage vector d ⁱ _c , calculate all the elements of the current actual damage vector d ⁱ .

                             式2

Formula 2

In formula 2, j =1,2,3,...,N.

Since the element value of the current actual damage vector d ⁱ represents the damage degree of the corresponding cable, which cables are damaged and the damage degree can be determined according to the current actual damage vector d ⁱ , that is, the health monitoring of the cable system in the cable structure is realized; If the value of an element of the current actual damage vector is 0, it means that the cable corresponding to this element is intact and has no damage; if its value is 100%, it means that the cable corresponding to this element has completely lost its bearing capacity; If its value is between 0 and 100%, it means that the cable has lost a corresponding proportion of its carrying capacity.

m. After obtaining the current nominal damage vector d ⁱ _c , establish the identification vector F ⁱ according to formula 3, and formula 4 gives the definition of the jth element of the identification vector F ⁱ ;

            式3

Formula 3

                              式4

Formula 4

In formula 4, the element F ⁱ _j is the jth element of the identification vector F ⁱ , D ⁱ _uj is the jth element of the nominal unit damage vector D ⁱ _u , and d ⁱ _cj is the jth element of the current nominal damage vector d ⁱ _c of the cable system j elements, they all represent the relevant information of the jth root cable. In formula 4, j =1, 2, 3,...,N.

n． If the elements of the identification vector F ⁱ are all 0, then go back to step g to continue this cycle; if the elements of the identification vector F ⁱ are not all 0, then enter the next step, namely step o.

o. Calculate each element d i + 1 _oj of the initial damage vector d i ^{+ 1} _o required for the next cycle, that is, ^{the i + 1th} cycle according to formula 5;

                               式5

Formula 5

In formula 5, D ⁱ _uj is the jth element of the nominal unit damage vector D ⁱ _u , d ⁱ _cj is the jth element of the current nominal damage vector d ⁱ _c of the cable system, and F ⁱ _j is the jth element of the identification vector F ⁱ elements. In formula 5, j =1, 2, 3,...,N.

p. On the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, let the health status of the cable be d ^{i + 1} _o and then update to obtain the mechanical calculation benchmark model A ⁱ required for the next cycle i+1 ⁺¹ , that is, the benchmark model for mechanical calculations has been updated;

q. Obtain the values of all the monitored quantities corresponding to the structure of the model A ⁱ⁺¹ through the calculation of the mechanical calculation benchmark model A ⁱ⁺¹ , and these values constitute the monitored values required for the next cycle i+1 The initial numerical vector C ⁱ⁺¹ _o of the quantity;

r. Establish the current mechanical calculation benchmark model A ^ti+1 _o of the cable structure required for the next cycle i+1, that is, take A ^ti+1 _o equal to A ⁱ⁺¹ ;

s. Establish the current cable structure support angular coordinate vector U ^{ti + 1} _o required for the next cycle, that is, the i+1th cycle, that is, take U ^{ti + 1} _o equal to U ^ti _o ;

t. Go back to step g and start the next cycle.

the

In step g, according to the measured support angle coordinate vector U ^ti of the current cable structure, the specific method for updating the current mechanical calculation benchmark model A ^ti _{o of} the cable structure and the current support angle coordinate vector U ^ti _o of the cable structure is as follows:

g1. After obtaining the measured support angular coordinate vector U ^ti of the current cable structure, compare U ^ti with U ^ti _o , if U ^ti is equal to U ^ti _o , there is no need to update A ^ti _o ;

g2. After obtaining the measured support angular coordinate vector U ^ti of the current cable structure, compare U ^ti and U ^ti _o , if U ^ti is not equal to U ^ti _o , then update A ^ti _o , the update method is: calculate U first The difference between ^ti and U _o , the difference between U ^ti and U _o is the current support angular displacement of the current cable structure support relative to the cable structure support when A _o is established, and the support angle is expressed by the current support angular displacement vector V Displacement. There is a one-to-one correspondence between the elements in the current support angular displacement vector V and the support angular displacement components. The value of an element in the current support angular displacement vector V corresponds to a specified support around a specified direction. Angular displacement; the method to update A ^ti _o is: on the basis of A _o , let the health status of the cable be the initial damage vector d ⁱ _o of the cable system, and then further impose the current support angular displacement constraint on the cable structure support in A _o , the value of the current support angular displacement constraint is taken from the value of the corresponding element in the current support angular displacement vector V , after applying the current support angular displacement constraint to the cable structure support in A _o , the final result is the updated current Mechanical calculation benchmark model A ^ti _o , when A ^ti _o is updated, all element values of U ^ti _o are also replaced by all element values of U ^ti , that is, U ^ti _o is updated, so that the U that correctly corresponds to A ^ti _o is obtained ^ti _o .

the

In step h, several mechanical calculations are performed on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, and the specific method of obtaining the change matrix ΔC ⁱ of the monitored quantity of damage per unit of the cable structure and the nominal unit damage vector D ⁱ _u through calculation for:

h1. At the beginning of the ith cycle, directly follow the methods listed in step h2 to step h4 to obtain the cable structure unit damage monitored variable matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u ; After A ^ti _o is updated, the cable structural unit damage monitored variable matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u must be obtained according to the methods listed in step h2 to step h4. If A ^ti _o is not updated in step g , then go directly to step i here for follow-up work;

h2. Carry out several mechanical calculations on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure. The number of calculations is numerically equal to the number of all cables, and there are N calculations for N cables. Each calculation assumes that there are only A cable adds unit damage on the basis of the original damage. The damaged cable in each calculation is different from the damaged cable in other calculations, and the unit damage value of the damaged cable can be different from other cables in each calculation. The unit damage value of , use the "nominal unit damage vector D ⁱ _u " to record the assumed unit damage of all cables, and get the current values of all the specified monitored quantities in the cable structure for each calculation, and get the current values of all the monitored quantities for each calculation The current value forms a "calculated current value vector of the monitored quantity"; when it is assumed that the jth root cable has unit damage, C ⁱ _tj can be used to represent the corresponding "calculated current value vector of the monitored quantity"; in this step, each When the element numbering of vector, should use same numbering rule with other vector among the present invention, can guarantee like this any one element in each vector in this step, with the same element of numbering in other vector, expressed same monitored quantity or Information about the same subject;

h3. Subtract the "monitored quantity's _calculated current value vector C ⁱ _tj " from each calculation to get a vector, and then divide each element of ^the vector by After the unit damage value assumed in this calculation, a "value change vector δC ⁱ _j of the monitored quantity" is obtained; if there are N cables, there are N "value change vectors of the monitored quantity";

h4. The N "monitored quantity change matrix ^{ΔC i " with N columns is composed in turn of the N "monitored quantity change vectors"; each column of the "monitored quantity change matrix ΔC i per} unit damage " corresponds to a "Numerical change vector of the monitored quantity"; the numbering rules of the columns of the "unit damage monitored quantity change matrix" are the same as the element numbering rules of the current nominal damage vector d ⁱ _c and the current actual damage vector d ⁱ .

the

Beneficial effects: the system and method disclosed in the present invention can very accurately monitor and evaluate the health status of the cable system (including all affected location and extent of damage). This is due to the difference between "the current numerical vector C ⁱ of the monitored quantity" and "the initial numerical vector C ⁱ _o of the monitored quantity", "the change matrix ΔC ⁱ of the monitored quantity with unit damage" and "the current nominal damage vector d ⁱ _c " The linear relationship is approximate, but in fact it is a nonlinear relationship, especially when there are many damaged cables or the damage degree is relatively large, the nonlinear characteristics of the relationship between the above-mentioned quantities are more obvious. In order to overcome this obstacle, the present invention Disclosed is a health monitoring method for approximating the nonlinear relationship with a linear relationship within a small area under the condition that the angular displacement of the support of the cable structure occurs. The present invention actually uses the method of segmentally approaching the nonlinear relationship with the linear relationship, and divides the large interval into small intervals. The linear relationship in each small interval is accurate enough, and the cable system obtained according to its judgment Health status is also reliable. Therefore, the system and method disclosed in the present invention is very beneficial for the effective health monitoring of the cable system in the case of angular displacement of the support of the cable structure.

Detailed ways

For the health monitoring of the cable system of the cable structure when there is angular displacement of the support, the invention discloses a system and method capable of reasonably and effectively monitoring the health status of each cable of the cable system of the cable structure. The following descriptions of embodiments of the invention are merely exemplary in nature, and are in no way intended to limit the application or uses of the invention.

The algorithm employed by the present invention is used to monitor the health status of the cable system in the cable structure in case of angular displacement of the cable structure support. During specific implementation, the following steps are one of various steps that may be taken.

Step 1: Determine the type, location and quantity of the quantity to be monitored, and number them. The specific process is:

Assuming that there are N cables, first determine the numbering rules of the cables, and number all the cables in the cable structure according to this rule, and the numbers will be used to generate vectors and matrices in subsequent steps.

Determine the designated measured points (that is, all designated points representing the angular displacement of the structure, with K designated points), and number all designated points; determine the measured straight line of each measured point (set each measured point with L specified straight line), number all specified measured straight lines; determine the measured angular coordinate components of each measured straight line (assuming that each measured straight line has H angular coordinate components), and number all measured angular coordinate components . The above numbers will also be used to generate vectors and matrices in subsequent steps. The "all monitored angle data of the structure" is described by the K specified points on the structure determined above, the L specified straight lines passing through each specified point, and the H angular coordinate components of each specified straight line. The change is the change of all specified angle coordinate components of all specified points and all specified lines. Each time there are M (M=K×L×H) angle coordinate component measured or calculated values to represent the angle information of the structure. K and M shall not be less than the number N of supporting cables. For convenience, the "monitored angle data of the structure" is referred to as "monitored quantity" in the present invention. At each specified point, only one angular coordinate of a specified straight line can be measured, for example, the angular coordinate of the surface normal of the specified point relative to the direction of gravitational acceleration is measured, which is actually an inclination measurement here.

The second step: use the non-destructive testing data of the cable and other data that can express the health state of the cable to establish the initial damage vector d ¹ _o of the cable system. If there is no non-destructive testing data of the cable and other data that can express the healthy state of the cable, or when the initial state of the structure can be considered as a non-damaged state, the value of each element of the vector d ¹ _o is 0.

Step 3: While establishing the initial damage vector d ¹ _o of the cable system, directly measure and calculate all the specified monitored quantities of the cable structure to form the "initial value vector C ¹ _o of the monitored quantities".

Step 4: While establishing the initial damage vector d ¹ _o of the cable system and the initial value vector C ¹ _o of the monitored quantity, mature measurement methods can be used for cable force measurement, strain measurement, angle measurement and space coordinate measurement. At the same time, the initial cable force of all cables of the cable structure and the initial geometric shape data of the cable structure (for cable-stayed bridges is the initial bridge type data) of the cable structure can be obtained directly or calculated after measurement. The initial geometric shape data of the cable structure can be all cables The spatial coordinate data of the end points plus the spatial coordinate data of a series of points on the structure, the purpose is to determine the geometric characteristics of the cable structure according to these coordinate data. For cable-stayed bridges, the initial geometric shape data can be the spatial coordinate data of all cable end points plus the spatial coordinate data of several points on both ends of the bridge, which is the so-called bridge type data.

Establish the initial mechanical calculation benchmark model A _o of the cable structure, establish the initial cable structure support angle coordinate vector U _o , and establish the mechanical calculation benchmark model A ¹ of the cable structure required at the beginning of the first cycle; The actual measurement data of the cable structure, the actual measurement data includes the cable structure shape data, the cable force data, the tie rod tension data, the cable structure support coordinate data, the cable structure support angular coordinate data, the cable structure modal data and other measured data. Test data and other data that can express the health status of the cable, based on the design drawing and as-built drawing, use mechanical methods to establish the initial mechanical calculation benchmark model A _o of the cable structure; if there is no actual measurement data of the cable structure when it is completed, then in The actual measurement of the cable structure is carried out before the health monitoring system is established, and the actual measurement data of the cable structure is also obtained. According to this data and the design drawing and as-built drawing of the cable structure, the initial mechanical calculation benchmark model A _o of the cable structure is also established by the mechanical method; no matter What method is used to obtain A _o , the cable structure calculation data based on A _o must be very close to its measured data, and the difference between them is generally not greater than 5%; the cable structure support angle coordinate data corresponding to A _o constitutes the initial cable structure Support angular coordinate vector U _o ; A _o and U _o are constant, and are only established at the beginning of the first cycle; the mechanical calculation benchmark model of the cable structure established at the beginning of the i-th cycle is denoted as A ⁱ , where i Indicates the number of cycles; in the application of the present invention, letter i only indicates the number of cycles, i.e. the i-th cycle, except where the letter i clearly indicates the step number in the application of the present invention; therefore the mechanical calculation benchmark of the cable structure established at the beginning of the first cycle The model is denoted as A ¹ , and A ¹ is equal to A _o in the present invention; _for the convenience of description, it is named "the current mechanical calculation benchmark model ^{A ti} _o of the cable structure". At the beginning of the cycle, A ^ti _o is equal to A ⁱ ; also for the convenience of description, it is named "cable structure measured support angle coordinate vector U ^ti ". In each cycle, the current data of the cable structure support angle coordinates are continuously measured. The current data of the structural support angular coordinates form the current cable structure measured support angular coordinate vector U ^ti , the elements of the vector U ^ti and the elements of the same position as the vector U _o represent the angular coordinates of the same support in the same direction; for the convenience of description, for In the i-th cycle, the current data of the cable structure support angular coordinates when A ^ti _o was updated last time is recorded as the current cable structure support angular coordinate vector U ^ti _o ; at the beginning of the first cycle, A ^t1 _o is equal to A ¹ , U ^t1 _o is equal to U _o .

Step 5: Install the hardware part of the cable structure health monitoring system. The hardware part includes at least: monitored quantity monitoring system (including angle measurement sensor, signal conditioner, etc.), cable structure support angular coordinate monitoring system (including angle measurement sensor, signal conditioner, etc.), signal (data) collector, computer and communication alarm equipment. Each monitored quantity and the support angle coordinates of each cable structure must be monitored by the monitoring system. The monitoring system transmits the monitored signal to the signal (data) collector; the signal is transmitted to the computer through the signal collector; The health monitoring software of the cable system responsible for running the cable structure includes recording the signal transmitted by the signal collector; when the cable is detected to be damaged, the computer controls the communication alarm device to alarm the monitoring personnel, the owner and (or) the designated personnel.

Step 6: Compile and install the cable system health monitoring system software of the cable structure on the monitoring computer. The software is run on every cycle, or the software is always running. This software will complete the monitoring, recording, control, storage, calculation, notification, alarm and other functions required by the task of the present invention "progressive health monitoring method based on angle monitoring cable system" (that is, this specific implementation All the work that can be done by computer in the method), and the health status report of the cable system can be generated periodically or by personnel operating the health monitoring system, and can also automatically notify or prompt the monitoring personnel according to the set conditions (such as damage reaching a certain value) Notify specific technicians to complete the necessary calculations.

Step 7: Start the cyclic operation from this step, which is recorded as the i-th cycle for the convenience of description, where i=1, 2, 3, 4, 5,.... The actual measurement (including angle measurement sensor, signal conditioner, etc.) obtains the current data of the angular coordinates of the cable structure supports, and all the current data of the angular coordinates of the cable structure supports form the vector U ^ti of the measured Seat angle coordinate vector U ^ti , if necessary, update the current cable structure mechanical calculation benchmark model A ^ti _o and the current cable structure support angle coordinate vector U ^ti _o . The specific method is:

After obtaining the measured support angle coordinate vector U ^ti of the current cable structure, compare U ^ti with U ^ti _o , if U ^ti is equal to U ^ti _o , there is no need to update A ^ti _o ;

After obtaining the actual measured support angle coordinate vector U ^ti of the current cable structure, compare U ^ti and U ^ti _o , if U ^ti is not equal to U ^ti _o , it is necessary to update A ^ti _o , the update method is: first calculate U ^ti and The difference of U _o , the difference between U ^ti and U _o is the current support angular displacement of the current cable structure support with respect to the cable structure support when A _o is established, and the current support angular displacement vector V is used to represent the support angular displacement, There is a one-to-one correspondence between the elements in the current support angular displacement vector V and the support angular displacement components. The value of an element in the current support angular displacement vector V corresponds to the angular displacement of a specified support around a specified direction ; The method to update A ^ti _o is: on the basis of A _o , let the health status of the cable be the initial damage vector d ⁱ _o of the cable system, and then further impose the current support angular displacement constraint on the cable structure support in A _o , the current The value of the support angular displacement constraint is taken from the value of the corresponding element in the current support angular displacement vector V. After applying the current support angular displacement constraint to the cable structure support in A _o , the final result is the updated current mechanical calculation In the benchmark model A ^ti _o , when A ^ti _o is updated, all element values of U ^ti _o are also replaced by all element values of U ^ti , that is, U ^ti _o is updated, so that U ^ti _o correctly corresponding to A ^ti _o is obtained .

Step 8: Carry out several mechanical calculations on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, and obtain the cable structure unit damage monitored quantity change matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u through calculation. The specific method is:

a. At the beginning of the ith cycle or when A ^ti _o is updated in the seventh step, directly follow the methods listed in step b to step d to obtain the cable structure unit damage monitored variable matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u ; at other moments, if A ^ti _o is not updated in step g, then directly turn to the ninth step here for follow-up work;

b. Carry out several mechanical calculations on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure. The number of calculations is numerically equal to the number of all cables. There are N calculations for N cables. Each calculation assumes that there are only A cable adds unit damage on the basis of the original damage. The damaged cable in each calculation is different from the damaged cable in other calculations, and the unit damage value of the damaged cable can be different from other cables in each calculation. The unit damage value of , use the "nominal unit damage vector D ⁱ _u " to record the assumed unit damage of all cables, and get the current values of all the specified monitored quantities in the cable structure for each calculation, and get the current values of all the monitored quantities for each calculation The current value forms a "calculated current value vector of the monitored quantity"; when it is assumed that the jth root cable has unit damage, C ⁱ _tj can be used to represent the corresponding "calculated current value vector of the monitored quantity"; in this step, each When the element numbering of vector, should use same numbering rule with other vector among the present invention, can guarantee like this any one element in each vector in this step, with the same element of numbering in other vector, expressed same monitored quantity or Information about the same object.

c . Subtract the "monitored quantity ^'s calculated current value vector C ⁱ _tj " from each calculation to get a vector _, and then divide each element of the vector by After the unit damage value assumed in this calculation, a "value change vector δC ⁱ _j of the monitored quantity" is obtained; if there are N cables, there are N "value change vectors δC ⁱ _j of the monitored quantity" ( j=1, 2,3,…,N ).

d. A "monitored quantity change matrix ΔC ⁱ per unit damage" with N columns is composed in turn of the N "value change vectors of the monitored quantity"; each column of the "monitored quantity change matrix ΔC ⁱ unit damage" corresponds to a "Numerical change vector of the monitored quantity"; the numbering rules of the columns of the "unit damage monitored quantity change matrix" are the same as the element numbering rules of the current nominal damage vector d ⁱ _c and the current actual damage vector d ⁱ .

When numbering the elements of each vector in this step and thereafter, the same numbering rule should be used with other vectors in the present invention, so that it can be guaranteed that any element in each vector in this step has the same numbering as in other vectors. Elements express the related information of the same monitored quantity or the same object.

Step 9: Establish linear relationship error vector e ⁱ and vector g ⁱ . Using the previous data ("initial value vector C ⁱ _o of the monitored quantity", "monitored quantity change matrix ΔC ⁱ per unit damage"), while performing each calculation in the eighth step, that is, in each calculation, it is assumed that the index There is only one cable in the system, while adding unit damage on the basis of the original damage, each calculation forms a damage vector d ⁱ _t , the number of elements in the damage vector d ⁱ _t is equal to the number of cables, and all the vector d ⁱ _t The value of only one of the elements is the unit damage value of the cable that is assumed to increase the unit damage in each calculation, the value of the other elements of d ⁱ _t is 0, and the number of the element that is not 0 is the same as that of the cable that is assumed to increase the unit damage The corresponding relationship between elements with the same number as other vectors is the same as that of the index; put C ⁱ _tj , C ⁱ _o , ΔC ⁱ , d ⁱ _t into formula (13) to obtain a linear relationship error vector e ⁱ , each calculation gets a linear relationship error vector e ⁱ ; if there are N cables, there are N calculations, and there are N linear relationship error vectors e ⁱ , and the N linear relationship error vectors e ⁱ are added together to obtain a vector, the new vector obtained after dividing each element of this vector by N is the final linear relationship error vector e ⁱ . The vector g ⁱ is equal to the final error vector e ⁱ . Save the vector g ⁱ on the computer hard disk running the health monitoring system software for use by the health monitoring system software.

Step 10: Save parameters such as "initial value vector C ⁱ _o of the monitored quantity" and "variation matrix ΔC ⁱ of the monitored quantity per unit damage" in the form of data files on the hard disk of the computer running the health monitoring system software. The current measured values of all specified monitored quantities of the cable structure are obtained through actual measurement, and form a "current value vector C ⁱ of the monitored quantities".

Step 11: According to "the current numerical vector C ⁱ of the monitored quantity" and "the initial numerical vector C ⁱ _o of the monitored quantity", "the change matrix ΔC ⁱ of the monitored quantity with unit damage" and "the current nominal damage vector d ⁱ _c ” exists an approximate linear relationship (Equation (9)), according to the multi-objective optimization algorithm to calculate the non-inferior solution of the current nominal damage vector d ⁱ _c of the cable system.

There are many kinds of multi-objective optimization algorithms that can be used, such as: multi-objective optimization based on genetic algorithm, multi-objective optimization based on artificial neural network, multi-objective optimization algorithm based on particle swarm, multi-objective optimization based on ant colony algorithm, constraint method (Constrain Method), Weighted Sum Method, Goal Attainment Method, etc. Since various multi-objective optimization algorithms are conventional algorithms, they can be easily implemented. This implementation step only uses the objective programming method as an example to give the process of solving the current nominal damage vector d ⁱ _c . The specific implementation process of other algorithms can be based on their specific The requirements of the algorithm are implemented in a similar fashion.

According to the objective programming method, Equation (9) can be transformed into the multi-objective optimization problem shown in Equation (21) and Equation (22). In Equation (21), γ ⁱ is a real number, R is a real number field, and the spatial region Ω limits The value range of each element of the vector d ⁱ _c (this embodiment requires that each element of the vector d ⁱ _c is not less than 0 and not greater than 1). Equation (21) means to find a real number γ ⁱ with the smallest absolute value, so that Equation (22) is satisfied. G(d ⁱ _c ) in formula (22) is defined by formula (23), and the product of weighted vector W ⁱ and γ ⁱ in formula (22) represents the relationship between G(d ⁱ _c ) and vector g ⁱ in formula (22) For the allowable deviation, the definition of g ⁱ can be found in formula (15), and its value will be calculated in the eighth step. The vector W ⁱ may be the same as the vector g ⁱ in actual calculation. The specific programming implementation of the goal programming method already has a general program that can be directly adopted. According to the goal programming method, the current nominal damage vector d ⁱ _c can be obtained .

                                        (21)

(twenty one)

                                       (22)

(twenty two)

                            (23)

(twenty three)

After obtaining the current nominal damage vector d ⁱ _c , each element of the current actual damage vector d ⁱ can be obtained according ^to formula (17). A solution to determine the location of the damaged cable and its extent of damage. If the value of an element of the current actual damage vector d ⁱ is 0, it means that the cable corresponding to this element is intact and has no damage; if its value is 100%, it means that the cable corresponding to this element has been damaged. Complete loss of bearing capacity; if its value is between 0 and 100%, it means that the cable has lost a corresponding proportion of its bearing capacity.

Step 12: After obtaining the current nominal damage vector d ⁱ _c in this cycle, that is, the i -th cycle, firstly, establish the identification vector F ⁱ according to formula (18) and formula (19) . If the elements of the identification vector F ⁱ are all 0, then go back to the seventh step to continue this cycle; if the elements of the identification vector F ⁱ are not all 0, then enter the next step, that is, the thirteenth step.

Step 13: Calculate and obtain each _element d i + 1 oj of the initial damage vector d ^{i + 1} _o required for the next cycle, ie, the i ^{+ 1th} cycle, according to formula (20).

Step 14: On the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, let the health status of the cable be the vector d ^{i + 1} _o calculated in the previous step, and obtain a new mechanical calculation benchmark model, that is, the next The mechanical calculation benchmark model A i ⁺¹ required for the (i+1th) cycle.

Step 15: Obtain the values of all the monitored quantities corresponding to the structure of the model A ⁱ⁺¹ through the calculation of the mechanical calculation benchmark model A ⁱ⁺¹ , and these values constitute the requirements for the next cycle, that is, the i+1th cycle The vector C ⁱ⁺¹ _{o of} is the initial numerical vector of the monitored quantity .

The sixteenth step: establish the current mechanical calculation benchmark model A ^ti+1 _o of the cable structure required for the next cycle i+1, that is, take A ^ti+1 _o equal to A ⁱ⁺¹ .

Step 17: Establish the current cable structure support angular coordinate vector U ^{ti + 1} _o required for the next cycle, that is, the i+1th cycle, that is, take U ^{ti + 1} _o to be equal to U ^ti _o .

Step 18: Go back to Step 7 and start the next cycle.

Claims (3)

1. The progressive health monitoring method of the cable system based on angle monitoring during a bearing angular displacement, it is characterized in that the method comprises: a. Assuming that there are N cables in total, first determine the numbering rules of the cables, and number all the cables in the cable structure according to this rule, and the numbers will be used to generate vectors and matrices in subsequent steps; b. Determine the specified measured point, and number all the specified points; determine the measured straight line of each measured point, and number all the specified measured straight lines; determine the measured angular coordinate component of each measured straight line, and give The numbers of all measured angular coordinate components. The above numbers will be used in subsequent steps to generate vectors and matrices. The "all monitored angular data of the structure" consists of all the above-mentioned measured angular coordinate components. For the sake of convenience, the "monitored angle data of the structure" is simply referred to as "monitored quantity" in the present invention. The number of measuring points shall not be less than the number of cables; the sum of the numbers of all measured angle coordinate components shall not be less than the number of cables; c. Establish the initial damage vector d ¹ _o of the cable system by using the nondestructive testing data of the cable and other data that can express the healthy state of the cable; if there is no nondestructive testing data of the cable and other data that can express the healthy state of the cable, the structure When the initial state is a non-damaged state, the value of each element of the vector d ¹ _o is 0; the superscript 1 of d ¹ _o in this step indicates the first cycle, and the expression method of the number of cycles is specifically explained in step f; d. While establishing the initial damage vector d ¹ _o of the cable system, directly measure and calculate all the specified monitored quantities of the cable structure to form the "initial value vector C ¹ _o of the monitored quantity"; in this step, the value of C ¹ _o The superscript 1 represents the first cycle, and the representation method of the number of cycles is specified in step f; e. While establishing the initial damage vector d ¹ _o of the cable system and the initial numerical vector C ¹ _o of the monitored quantity, the initial cable force data of all cables of the cable structure are obtained through actual measurement, and the initial geometric data of the cable structure are obtained through actual measurement; f. Establish the initial mechanical calculation benchmark model A _o of the cable structure, establish the initial cable structure support angle coordinate vector U _o , and establish the mechanical calculation benchmark model A ¹ of the cable structure required at the beginning of the first cycle; in this step, A ¹ The superscript 1 in the indicates the first cycle; according to the measured data of the cable structure when the cable structure is completed, the measured data includes the shape data of the cable structure, the data of the The measured data such as seat angle coordinate data, cable structure modal data, cable non-destructive testing data and other data that can express the health status of the cable, based on the design drawing and as-built drawing, use the mechanical method to establish the initial mechanical calculation benchmark model A _o of the cable structure. ; If there is no actual measurement data of the structure when the cable structure is completed, then the actual measurement of the cable structure is carried out before the establishment of the health monitoring system, and the actual measurement data of the cable structure is also obtained. Also use the mechanical method to establish the initial mechanical calculation benchmark model A _o of the cable structure; no matter what method is used to obtain A _o , the calculated data of the cable structure based on A _o must be very close to the measured data, and the difference between them shall not be greater than 5%; The angular coordinate data of the cable structure support corresponding to A _o constitutes the initial angular coordinate vector U _o of the cable structure support; A _o and U _o are invariable and are only established at the beginning of the first cycle; at the beginning of the i-th cycle The mechanical calculation reference model of the established cable structure is denoted as A ⁱ , where i represents the number of cycles; in the application of the present invention, the letter i only represents the number of cycles, i.e. the i-th cycle, except where the number of steps is clearly indicated. ; Therefore, the mechanical calculation benchmark model of the cable structure established at the beginning of the first cycle is recorded as A ¹ , and A ¹ is equal to A _o in the present invention; for the convenience of description, it is named "the current mechanical calculation benchmark model of the cable structure A ^ti _o ", In each cycle, A ^ti _o will be continuously updated according to the needs. At the beginning of each cycle, A ^ti _o is equal to A ⁱ ; also for the convenience of description, it is named "cable structure measured support angle coordinate vector U ^ti ", and in each cycle In this method, the current data of the support angle coordinates of the cable structure are continuously measured, and all the current data of the support angle coordinates of the cable structure _form the current cable structure measured support angle coordinate vector U ^ti , and the elements of the vector U ^ti are represented by the elements at the same position as the vector U o The angular coordinates of the same support in the same direction; for the convenience of description, for the i-th cycle, the current data of the angular coordinates of the cable structure support when A ^ti _o is updated last time is recorded as the current cable structure support angular coordinate vector U ^ti _o ; At the beginning of the first cycle, A ^t1 _o is equal to A ¹ , and U ^t1 _o is equal to U _o ; g. Obtain the current data of the support angle coordinates of the cable structure through actual measurement. All the current data of the support angle coordinates of the cable structure form the current cable structure measured support angle coordinate vector U ^ti . According to the current cable structure measured support angle coordinate vector U ^ti , if necessary Timely update the current mechanical calculation benchmark model A ^ti _o of the cable structure and the current support angular coordinate vector U ^ti _o of the cable structure; h. Carry out several mechanical calculations on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, and obtain the cable structure unit damage monitored quantity change matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u through calculation; i. The current measured values of all the specified monitored quantities of the cable structure are obtained through actual measurement, and form the "current value vector C ⁱ of the monitored quantities". When numbering the elements of all vectors that appear in this step and before this step, the same numbering rule should be used, so as to ensure that the elements with the same number in each vector that appeared in this step and before this step represent the corresponding value of the same monitored quantity. related information defined in the vector to which the element belongs; j. Define the current nominal damage vector d ⁱ _c and the current actual damage vector d ⁱ of the cable system. The number of elements in the damage vector is equal to the number of cables. There is a one-to-one correspondence between the elements of the damage vector and the cables. The value of the elements of the damage vector Represents the degree of damage or health status of the corresponding cable; k. According to the relationship between "the current numerical vector C ⁱ of the monitored quantity" and "the initial numerical vector C ⁱ _o of the monitored quantity", "the change matrix ΔC ⁱ of the monitored quantity with unit damage" and "the current nominal damage vector d ⁱ _c " Existing approximate linear relationship, the approximate linear relationship can be expressed as formula 1, in formula 1, other quantities except d ⁱ _c are known, and the current nominal damage vector d ⁱ _c can be calculated by solving formula 1;

Formula 1 l. Using the relationship between the current actual damage vector d i expressed in formula 2 ^and the elements of the initial damage vector d ⁱ _o and the current nominal damage vector d ⁱ _c , calculate all the elements of the current actual damage vector d ⁱ .

Formula 2 In formula 2, j =1,2,3,...,N. Since the element value of the current actual damage vector d ⁱ represents the damage degree of the corresponding cable, which cables are damaged and the damage degree can be determined according to the current actual damage vector d ⁱ , that is, the health monitoring of the cable system in the cable structure is realized; If the value of an element of the current actual damage vector is 0, it means that the cable corresponding to this element is intact and has no damage; if its value is 100%, it means that the cable corresponding to this element has completely lost its bearing capacity; If its value is between 0 and 100%, it means that the cable has lost a corresponding proportion of its carrying capacity. m. After obtaining the current nominal damage vector d ⁱ _c , establish the identification vector F ⁱ according to formula 3, and formula 4 gives the definition of the jth element of the identification vector F ⁱ ;

Formula 3

Formula 4 In formula 4, the element F ⁱ _j is the jth element of the identification vector F ⁱ , D ⁱ _uj is the jth element of the nominal unit damage vector D ⁱ _u , and d ⁱ _cj is the jth element of the current nominal damage vector d ⁱ _c of the cable system j elements, they all represent the relevant information of the jth root cable. In formula 4, j =1, 2, 3,...,N. n． If the elements of the identification vector F ⁱ are all 0, then go back to step g to continue this cycle; if the elements of the identification vector F ⁱ are not all 0, then enter the next step, namely step o. o. Calculate each element d i + 1 _oj of the initial damage vector d i ^{+ 1} _o required for the next cycle, that is, ^{the i + 1th} cycle according to formula 5;

Formula 5 In formula 5, D ⁱ _uj is the jth element of the nominal unit damage vector D ⁱ _u , d ⁱ _cj is the jth element of the current nominal damage vector d ⁱ _c of the cable system, and F ⁱ _j is the jth element of the identification vector F ⁱ elements. In formula 5, j =1, 2, 3,...,N. p. On the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure, let the health status of the cable be d ^{i + 1} _o and then update to obtain the mechanical calculation benchmark model A ⁱ required for the next cycle i+1 ⁺¹ , that is, the benchmark model for mechanical calculations has been updated; q. Obtain the values of all the monitored quantities corresponding to the structure of the model A ⁱ⁺¹ through the calculation of the mechanical calculation benchmark model A ⁱ⁺¹ , and these values constitute the monitored values required for the next cycle i+1 The initial numerical vector C ⁱ⁺¹ _o of the quantity; r. Establish the current mechanical calculation benchmark model A ^ti+1 _o of the cable structure required for the next cycle i+1, that is, take A ^ti+1 _o equal to A ⁱ⁺¹ ; s. Establish the current cable structure support angular coordinate vector U ^{ti + 1} _o required for the next cycle, that is, the i+1th cycle, that is, take U ^{ti + 1} _o equal to U ^ti _o ; t. Go back to step g and start the next cycle. 2. The progressive health monitoring method of the cable system based on angle monitoring during the angular displacement of the bearing according to claim 1, wherein in step g, according to the current cable structure measured bearing angular coordinate vector U ^ti , in The specific method for updating the current mechanical calculation benchmark model A ^ti _o of the cable structure and the current support angle coordinate vector U ^ti _o of the cable structure when necessary is as follows: g1. After obtaining the measured support angular coordinate vector U ^ti of the current cable structure, compare U ^ti with U ^ti _o , if U ^ti is equal to U ^ti _o , there is no need to update A ^ti _o ; g2. After obtaining the measured support angular coordinate vector U ^ti of the current cable structure, compare U ^ti and U ^ti _o , if U ^ti is not equal to U ^ti _o , then update A ^ti _o , the update method is: calculate U first The difference between ^ti and U _o , the difference between U ^ti and U _o is the current support angular displacement of the current cable structure support relative to the cable structure support when A _o is established, and the support angle is expressed by the current support angular displacement vector V Displacement. There is a one-to-one correspondence between the elements in the current support angular displacement vector V and the support angular displacement components. The value of an element in the current support angular displacement vector V corresponds to a specified support around a specified direction. Angular displacement; the method to update A ^ti _o is: on the basis of A _o , let the health status of the cable be the initial damage vector d ⁱ _o of the cable system, and then further impose the current support angular displacement constraint on the cable structure support in A _o , the value of the current support angular displacement constraint is taken from the value of the corresponding element in the current support angular displacement vector V , after applying the current support angular displacement constraint to the cable structure support in A _o , the final result is the updated current Mechanical calculation benchmark model A ^ti _o , when A ^ti _o is updated, all element values of U ^ti _o are also replaced by all element values of U ^ti , that is, U ^ti _o is updated, so that the U that correctly corresponds to A ^ti _o is obtained ^ti _o . 3. The progressive health monitoring method of the cable system based on angle monitoring when the bearing angular displacement according to claim 1 is characterized in that in step h, on the basis of the current mechanical calculation reference model A ^ti _o of the cable structure Carry out several mechanical calculations, and the specific method to obtain the change matrix ΔC ⁱ of the monitored quantity of cable structure unit damage and the nominal unit damage vector D ⁱ _u through calculation is as follows: h1. At the beginning of the ith cycle, directly follow the methods listed in step h2 to step h4 to obtain the cable structure unit damage monitored variable matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u ; After A ^ti _o is updated, the cable structural unit damage monitored variable matrix ΔC ⁱ and the nominal unit damage vector D ⁱ _u must be obtained according to the methods listed in step h2 to step h4. If A ^ti _o is not updated in step g , then go directly to step i here for follow-up work; h2. Carry out several mechanical calculations on the basis of the current mechanical calculation benchmark model A ^ti _o of the cable structure. The number of calculations is numerically equal to the number of all cables, and there are N calculations for N cables. Each calculation assumes that there are only A cable adds unit damage on the basis of the original damage. The damaged cable in each calculation is different from the damaged cable in other calculations, and the unit damage value of the damaged cable can be different from other cables in each calculation. The unit damage value of , use the "nominal unit damage vector D ⁱ _u " to record the assumed unit damage of all cables, and get the current values of all the specified monitored quantities in the cable structure for each calculation, and get the current values of all the monitored quantities for each calculation The current value forms a "calculated current value vector of the monitored quantity"; when it is assumed that the jth root cable has unit damage, C ⁱ _tj can be used to represent the corresponding "calculated current value vector of the monitored quantity"; in this step, each When the element numbering of vector, should use same numbering rule with other vector among the present invention, can guarantee like this any one element in each vector in this step, with the same element of numbering in other vector, expressed same monitored quantity or Information about the same subject; h3. Subtract the "monitored quantity's _calculated current value vector C ⁱ _tj " from each calculation to get a vector, and then divide each element of ^the vector by After the unit damage value assumed in this calculation, a "value change vector δC ⁱ _j of the monitored quantity" is obtained; if there are N cables, there are N "value change vectors of the monitored quantity"; h4. The N "monitored quantity change matrix ^{ΔC i " with N columns is composed in turn of the N "monitored quantity change vectors"; each column of the "monitored quantity change matrix ΔC i per} unit damage " corresponds to a "Numerical change vector of the monitored quantity"; the numbering rules of the columns of the "unit damage monitored quantity change matrix" are the same as the element numbering rules of the current nominal damage vector d ⁱ _c and the current actual damage vector d ⁱ .