Design method of concentric circular ring sparse array circular polarization phased array antenna of random rotation array

Abstract

The invention discloses a design method of a random rotation array concentric circular ring sparse array circular polarization phased array antenna. The design method comprises the following steps: establishing a sparse concentric circular ring array optimization model and parameterizing the sparse concentric circular ring array optimization model; optimizing parameterized parameters through a non-dominated sorting multi-target differential evolution algorithm by using an antenna array overall directional diagram of a sparse concentric circular ring array optimization model and taking requirements of indexes including antenna sidelobe suppression, beam width and antenna unit spacing as targets; randomly rotating the antenna units in the four rotationally symmetrical sub-arrays; and obtaining a segmentation line. According to the invention, a sparse concentric circular ring array arrangement mode, a non-dominated sorting multi-target differential evolution algorithm and random rotation array formation are adopted, and an integrated assembly array is adopted, so that only one type of model required to be manufactured by the whole array is ensured, the number of gaps between the sub-arrays is small, and due to the fact that the sub-arrays are tightly attached to each other and the gaps are very small, the manufacturing efficiency of the array is greatly improved. Therefore, parasitic radiation caused by discontinuity of the gap is weakened, and the performance of the antenna array is improved.

Description

Technical Field

[0001] The invention relates to a phased array antenna design method, in particular to a randomly rotated concentric ring sparse array circularly polarized phased array antenna design method. Background Art

[0002] Due to the precious space and tight payload of the satellite-borne communication phased array antenna, the array size, weight, power consumption and manufacturing cost of the antenna design are all subject to high restrictions. The pursuit of using fewer antenna units to achieve the required indicators has become the mainstream trend in satellite antenna design. At present, the communication method between satellite-borne phased array antennas and ground equipment is mostly frequency division full-duplex system, that is, the satellite platform has two phased array antennas for receiving and transmitting. Inevitably, the transmitting array antenna will interfere with the receiving array antenna.

[0003] Phased array antennas are usually arranged in a regular pattern, which has the following disadvantages:

[0004] (1) In the regular array arrangement, the array elements are densely arranged, and the mutual coupling between the array elements is higher, which has an adverse effect on the active standing waves of the array elements. Sometimes, a decoupling structure needs to be added, which increases the design difficulty and manufacturing cost;

[0005] (2) With the same gain requirement, a regular array requires a larger number of array elements, resulting in higher cost and weight;

[0006] (3) Multi-beam phased arrays all use an AOB-like array architecture, that is, the power amplifier, low-noise amplifier, and amplitude-phase multifunctional RF chips are all surface mounted on the PCB board. If a regular arrangement design is used, in order to avoid the appearance of grating lobes, the transmitting array element spacing dt should be less than (1-1 / M) / (1+sinθ0)*λh=7.35mm, and the receiving array element spacing dr should be less than (1-1 / N) / (1+sinθ0)*λh=5mm. At this spacing, the device is difficult to surface mount and integrate, and the circuit is difficult to layout;

[0007] (4) When the array is arranged regularly, the heat source distribution is more dense and the space is more limited, which makes the heat dissipation design more difficult and is not conducive to the heat dissipation of the phased array;

[0008] (5) The array surface of the regularly arranged antenna array is composed of a large number of separately processed small sub-arrays, leaving gaps at the junctions between the sub-arrays. When this array surface structure is applied to circularly polarized array antennas, gap parasitic radiation may occur at certain frequencies within the working frequency band. Since parasitic radiation is uncontrolled and takes away part of the energy that should be normal electromagnetic radiation, it leads to reduced array gain, deterioration of sidelobe level and axial ratio;

[0009] In order to improve the circular polarization performance of circularly polarized array antennas in large-angle scanning and broadband satellite communication applications, the traditional array method uses sequential rotation technology. By arranging the array elements in a small subarray in a certain order to reduce the axial ratio of the beam, the circular polarization performance of a large array antenna composed of several subarrays can be effectively improved. However, when the beam is scanned, a high-level sidelobe will appear in the visible area, causing the sidelobe level to deteriorate and fail to meet the index requirements.

[0010] Based on the above defects of regular array, sparse layout will become a better choice. However, while the existing sparse array layout solution solves the above shortcomings, it will also cause new problems:

[0011] (1) There are more parameters controlling the arrangement of array elements and more requirements on sidelobe levels, so the array design is a multi-variable + multi-objective optimization problem. The solution of multi-variable + multi-objective optimization problem itself is very challenging;

[0012] (2) The sidelobe level of the array deteriorates, especially when a high-level sidelobe enters the visible area from the non-visible area during scanning, causing the sidelobe level to fail to meet the index requirements;

[0013] (3) The sequential rotation array method used by conventional arrays will no longer be usable (and it also has the aforementioned high sidelobe problem). How to reduce the beam axial ratio will become a new problem.

[0014] (4) There are many types of antenna boards, which increases the design workload and processing costs. In addition, it also brings inconvenience to material management and product assembly. Summary of the invention

[0015] The object of the present invention is to overcome the deficiencies of the prior art and to provide a design method for a randomly rotated concentric ring sparse array circularly polarized phased array antenna.

[0016] The objective of the present invention is achieved through the following technical solutions:

[0017] A first aspect of the present invention provides a design method for a randomly rotated concentric ring sparse array circularly polarized phased array antenna, comprising the following steps:

[0018] Establishing a sparse concentric circular array optimization model and parameterizing it; the sparse concentric circular array optimization model includes four rotationally symmetric subarrays, each subarray includes a plurality of antenna units, and the parameterized parameters include a radius of 1 to N circles, a number of antenna units of 1 to N circles, and a starting angle of 1 to N circles;

[0019] Using the overall antenna array pattern of the sparse concentric ring array optimization model, taking the requirements of indicators including antenna sidelobe suppression, beam width, and antenna unit spacing as the target, the parameterized parameters are optimized through a non-dominated sorting multi-objective differential evolution algorithm;

[0020] The antenna elements in the four rotationally symmetric subarrays were randomly rotated, and the simulation confirmed that the antenna array pattern met expectations;

[0021] The boundary between two sub-arrays is obtained, and the array is divided with a broken line as the starting point at the center of the array circle, so that the broken line passes through the center of the closest edges of adjacent units on both sides of the boundary and extends to the end of the array extension to obtain a dividing line; the dividing line is rotated 90° around the center of the circle and replicated three times.

[0022] Furthermore, the establishment of the sparse concentric ring array optimization model and parameterization includes:

[0023] Calculate the required aperture size and number of antenna units based on gain requirements;

[0024] The number of antenna units is an integer multiple of 4, and these antenna units respectively form four rotationally symmetric sub-arrays;

[0025] Select the 90° sector to allocate array element positions, determine the number of circles according to the number of antenna elements, and allocate the number of elements per circle according to the arc length of each circle;

[0026] The array elements in each circle are evenly arranged on a 90° arc, and the angle difference between the array elements relative to the center of the array surface is calculated;

[0027] Assign a starting angle to the starting element of each circle of elements, where the starting angle is greater than 0 and not greater than the angle difference between the elements;

[0028] After rotating the obtained quarter sub-array 90° around the center of the array and the surface four times, the complete antenna array is obtained. At this time, the position of each array element on the array surface is uniquely determined;

[0029] The antenna array is divided into N r circles, a total of 3N parameters are required to constrain the position of each element in the antenna array, which are: the radius of 1 to N circles, namely R1, R2, ... R N , the number of antenna units in 1 to N circles is N1, N2, ... N N , the starting angle of 1 to N circles is Complete the parameterization of the array element positions in the concentric ring array.

[0030] Furthermore, the overall antenna array pattern of the sparse concentric ring array optimization model includes:

[0031] Determine the fluctuation range of the parameters;

[0032] The unit radiation pattern is obtained by simulating the antenna unit;

[0033] The overall radiation pattern of the synthesized antenna array is calculated by using the coordinates and radiation patterns of the antenna units in the array:

[0034]

[0035] Where EF is the array element pattern; N r The number of concentric ring arrays; N m is the number of array elements on the nth circle and is an integer multiple of 4 so as to divide the array into four rotationally symmetric regions; w n is the vector weight value of the nth circle array element; k is the wave number, and k = 2π / λ; r n is the radius of the nth circle; is the sampling angle in S1 space, and is the beam pointing angle;

[0036] beta(n,m) is the rotation angle of the mth unit on the nth circle relative to the x-axis. The angle difference between each unit is equal, and there is Where c is a constant, For variables, c and Determines the additional angular rotation of the unit on each circle;

[0037] For the array layout design, no amplitude and phase weighting is applied to avoid loss of transmit EIRP or receive G / T. Therefore, for all w n =1, so we have

[0038]

[0039] Furthermore, the target of the requirements of indicators including antenna sidelobe suppression, beam width, and antenna unit spacing includes:

[0040] First, the number of array elements on each circle is given, and then the radius r is set while satisfying the beam width and being greater than the minimum spacing. n and Optimize; Since the index requires different requirements for the sidelobe level when the beam is scanned to different angles, this problem is a multi-objective optimization problem and is expressed as follows:

[0041]

[0042]

[0043] St.HPBW s ≥3.5°

[0044] Δd≥Δd s =9.5mm

[0045] in, is the beam pointing angle The maximum sidelobe level required when the interference avoidance mode and non-interference avoidance mode There is a difference; Peak sidelobe level calculated for an individual pattern; HPBW s is the required minimum beam width; Δd is the spacing between adjacent array elements; Δd s is the minimum required spacing between adjacent elements.

[0046] Further, the multi-objective differential evolution algorithm for non-dominated sorting optimizes the parameterized parameters, including:

[0047] a. Parameter settings: population size, maximum execution number of generations, crossover constant, number of variables;

[0048] b. Randomly generate the initial population according to the following expression:

[0049]

[0050] c. Randomly select different individuals in the population and repeat the DE / rand / 1 mutation operation

[0051]

[0052] And get the population pop m ; Scaling factor F i,j It is not a constant. For each gene point variation, it is a random number between [0,1].

[0053] d. Population m Perform a two-point crossover operation and obtain the population pop c ;

[0054] e. Evaluate population pop c The experimental individuals and the target individual x i The fitness value of

[0055] f. Pop the population m and pop c Merge and perform non-dominated sorting;

[0056] g. Calculate the congestion and sort in descending order;

[0057] h. Select the first PS individuals as the next generation population, and re-perform non-dominated sorting, crowding calculation and descending order;

[0058] Repeat steps c to h above until the required number of generations is reached, and then save the results.

[0059] Furthermore, the calculation of the fitness value includes:

[0060] Calculate the array point corresponding to the parameter and the minimum distance D0 between the units;

[0061] Calculate the array pattern and main lobe half-power beam width W0;

[0062] Normalize the directional pattern, remove the main lobe part, and screen out all unqualified side lobe points with exceeding values ​​Pi and their total number Np according to the side lobe requirements;

[0063] The fitness calculation formula is as follows:

[0064]

[0065] D lim With W lim It is the minimum unit spacing and half-power bandwidth requirement. a1 and a2 are coefficients greater than 1, and b1 is a coefficient less than 1 and greater than 0. It is selected according to the situation.

[0066] Furthermore, the non-dominated sorting includes:

[0067] For each individual in the population, calculate the number of dominated individuals Ni and the number of dominated individuals Si; the domination means that all indicators of an individual A are not weaker than those of an individual B and there is an indicator that is better than that of individual B, that is, A dominates B; the indicators include bandwidth, minimum unit spacing and side lobes;

[0068] Find all individuals in the population whose Ni is 0, store them in set F1, and assign a uniform non-dominated ordinal number Ti to all individuals;

[0069] Subtract 1 from the remaining individuals S1 that are dominated by the set F1;

[0070] Repeat the above steps until all individuals are stratified into values ​​F1 to FN.

[0071] Furthermore, the congestion degree calculation includes:

[0072] According to the N parameters of each individual, place it in N-dimensional space;

[0073] Calculate the crowding operator Ri, where Ri is the distance between an individual and its nearest individual.

[0074] Furthermore, the descending order arrangement includes:

[0075] All individuals are given a non-dominated ordinal number and a crowding operator. When two individuals have different non-dominated rankings, the one with the smaller ranking number is taken. If the non-dominated rankings are the same, the less crowded individual, i.e. the one with a larger Ri, is taken, and the order is ranked in this order.

[0076] Furthermore, the antenna elements in the four rotationally symmetric sub-arrays are randomly rotated, and simulation confirms that the antenna array pattern meets expectations, including:

[0077] Obtain the coordinate position of the array element in the optimized 1 / 4 sub-array;

[0078] Use electromagnetic simulation software to paste the antenna unit model on each point in turn and number them;

[0079] Generate a random number array with a length equal to the number of units in the subarray, ranging from 0° to 360°;

[0080] Rotate the antenna units along the unit center axis in sequence according to the random angles corresponding to the numbers in the array;

[0081] Then the 1 / 4 subarray is rotated 90° around the array center axis and replicated three times to obtain the model of all units;

[0082] Complete the gaps between unit models, improve the conditions of materials, boundary conditions, and feed amplitude and phase, simulate the array, and confirm whether the antenna array radiation pattern after rotating at a random angle meets expectations.

[0083] The beneficial effects of the present invention are:

[0084] In an exemplary embodiment of the present invention, a sparse concentric ring array is used, and the spacing between array elements is larger than that in a regular array, so that the mutual coupling between units is reduced, the active standing wave at the port is lower, and no additional decoupling structure is required;

[0085] The multi-objective differential evolution algorithm of non-dominated sorting can optimize multiple optimization target parameters at the same time, which is more efficient and can meet the complex index requirements in the field of array computing.

[0086] The random rotation array is used to break the periodicity of the structure, and the energy of the periodically introduced high-level side lobes is evenly distributed to other side lobes, avoiding the deterioration of side lobes caused by energy concentration.

[0087] The use of an integrated assembly array not only ensures that only one version of the entire array needs to be produced, but also reduces the number of gaps between sub-arrays. Because the sub-arrays fit closely together, the gaps are also small, thus reducing the parasitic radiation caused by discontinuities at the gaps and improving the performance of the antenna array. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1A flowchart of a design method for a randomly rotated concentric ring sparse array circularly polarized phased array antenna provided in an exemplary embodiment of the present invention;

[0089] Figure 2 A schematic diagram of the layout of a concentric ring array provided in an exemplary embodiment of the present invention;

[0090] Figure 3 A flowchart of a multi-objective differential evolution algorithm for non-dominated sorting provided in an exemplary embodiment of the present invention;

[0091] Figure 4 A schematic diagram of a face segmentation method provided in an exemplary embodiment of the present invention;

[0092] Figure 5 A schematic diagram of a sequential rotation array provided in the prior art;

[0093] Figure 6 A schematic diagram of a random rotation array provided in an exemplary embodiment of the present invention;

[0094] Figure 7 It is a sequential rotation array scanning pattern provided in the prior art;

[0095] Figure 8 This is a random rotation array scanning direction pattern provided in an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0096] The technical solution of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0097] In the description of the present invention, it should be noted that the directions or positional relationships indicated by "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. are directions or positional relationships based on the drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0098] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, "installation", "connection" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0099] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0100] See also Figure 1 , Figure 1 A flow chart of a design method for a randomly rotated concentric ring sparse array circularly polarized phased array antenna provided in an exemplary embodiment of the present invention is shown, comprising the following steps:

[0101] Establishing a sparse concentric circular array optimization model and parameterizing it; the sparse concentric circular array optimization model includes four rotationally symmetric subarrays, each subarray includes a plurality of antenna units, and the parameterized parameters include a radius of 1 to N circles, a number of antenna units of 1 to N circles, and a starting angle of 1 to N circles;

[0102] Using the overall antenna array pattern of the sparse concentric ring array optimization model, taking the requirements of indicators including antenna sidelobe suppression, beam width, and antenna unit spacing as the target, the parameterized parameters are optimized through a non-dominated sorting multi-objective differential evolution algorithm;

[0103] The antenna elements in the four rotationally symmetric subarrays were randomly rotated, and the simulation confirmed that the antenna array pattern met expectations;

[0104] The boundary between two sub-arrays is obtained, and the array is divided with a broken line as the starting point at the center of the array circle, so that the broken line passes through the center of the closest edges of adjacent units on both sides of the boundary and extends to the end of the array extension to obtain a dividing line; the dividing line is rotated 90° around the center of the circle and replicated three times.

[0105] Specifically, in this exemplary embodiment, compared with the prior art, it has the following differences and advantages:

[0106] (1) The present invention adopts a sparse array method. The principle of the sparse array is to change the radius and rotation angle of each concentric circle so that the distribution pattern of array elements changes to obtain better results. The number of array elements remains unchanged but the position is changed.

[0107] (1-1) The present invention adopts a sparse concentric ring array arrangement, and the spacing between array elements is larger than that in a regular array arrangement. Therefore, the mutual coupling between units is reduced, the active standing wave at the port is lower, and no additional decoupling structure is required.

[0108] (1-2) Under the condition of the same array gain, the number of array elements in a sparse array is less than that in a regular array. This can reduce processing costs, design difficulty, and weight while maintaining good performance.

[0109] (1-3) The sparse array has more space reserved for other intelligent patrol systems and method devices under the AOB architecture with strong network security protection requirements, such as PA, LNA and amplitude and phase multifunctional chips, and the layout design difficulty and processing technology difficulty are reduced.

[0110] (1-4) Due to the increase in spacing, the number and concentration of heat sources are reduced, and there is more design space for heat dissipation.

[0111] (2) The multi-objective differential evolution algorithm for non-dominated sorting adopted by the present invention:

[0112] (2-1) This optimization algorithm can optimize multiple optimization target parameters simultaneously, which is more efficient and can meet the complex index requirements in the field of array computing;

[0113] (2-2) In the absence of prior knowledge, the algorithm will not fall into the local optimal solution, the solution is more accurate, and the problem of layout rigidity is avoided.

[0114] (3) The advantages of the random rotation array method used in the present invention are as follows:

[0115] (3-1) Improve the axial ratio and gain drop during beam scanning;

[0116] (3-2) Compared with the sequential rotation array, the random rotation array scheme breaks the periodicity of the structure and evenly distributes the energy of the periodically introduced high-level side lobes to other side lobes, thus avoiding the deterioration of side lobes caused by energy concentration.

[0117] Fixed-angle rotation array has certain rules, which will lead to the energy in the side lobes being concentrated according to the rules, making some side lobes stand out from the crowd. Random rotation array has no rules at all, the energy in the side lobes is evenly distributed, there is no prominent side lobe, which is more conducive to the suppression of side lobes

[0118] (3-3) Compared with sequential rotation arrays, random rotation arrays have a wider range of applications and can also be used in irregular arrays such as the concentric ring sparse array in the present invention.

[0119] (4) The advantages of the integrated assembly array of the present invention are as follows:

[0120] (4-1) Integrated assembly means that the array surface is divided into four identical rotationally symmetrical sub-arrays according to the layout characteristics. Each sub-array is a complete multi-layer PCB antenna array, so the length and number of gaps are greatly reduced, minimizing the impact on the antenna's reception or radiation performance.

[0121] (4-2) The types of PCB boards required for antenna array production are reduced, which reduces the cost, time and material management difficulty of design and processing;

[0122] (4-3) The number of array sub-arrays is smaller, which reduces the difficulty and time of assembly.

[0123] (4-4) The cutting method of the present application not only ensures that only one type of template is required for the entire array, but also reduces the number of gaps between sub-arrays. Furthermore, because the sub-arrays fit closely together and the gaps are very small, the parasitic radiation caused by the discontinuity at the gaps is reduced, thereby improving the performance of the antenna array.

[0124] The following content will elaborate on the specific implementation methods of each step:

[0125] More preferably, in an exemplary embodiment, the step of establishing a sparse concentric circular array optimization model and parameterizing the model includes:

[0126] Calculate the required aperture size and number of antenna units based on gain requirements;

[0127] The number of antenna units is an integer multiple of 4, and these antenna units respectively form four rotationally symmetric sub-arrays;

[0128] Select the 90° sector to allocate array element positions, determine the number of circles according to the number of antenna elements, and allocate the number of elements in each circle according to the arc length of each circle. When allocating, pay attention to the radius difference between two adjacent circles not being too small and the number of array elements in each circle not being too large, so as to ensure that the elements do not conflict with each other in physical space.

[0129] The array elements in each circle are evenly arranged on a 90° arc, and the angle difference between the array elements relative to the center of the array surface is calculated;

[0130] Assign a starting angle to the starting element of each circle of elements, where the starting angle is greater than 0 and not greater than the angle difference between the elements;

[0131] After rotating the obtained quarter sub-array 90° around the center of the array and the surface four times, the complete antenna array is obtained. At this time, the position of each array element on the array surface is uniquely determined;

[0132] The antenna array is divided into N r circles, a total of 3N parameters are required to constrain the position of each element in the antenna array, which are: the radius of 1 to N circles, namely R1, R2, ... R N, the number of antenna units in 1 to N circles is N1, N2, ... N N , the starting angles of 1 to N circles are φ1, φ2, …φ N , completing the parameterization of the array element positions in the concentric ring array.

[0133] More preferably, in an exemplary embodiment, the overall antenna array pattern of the sparse concentric ring array optimization model includes:

[0134] Determine the fluctuation range of the parameters;

[0135] The unit radiation pattern is obtained by simulating the antenna unit;

[0136] The overall antenna array pattern is calculated by using the coordinates of the antenna units in the array and the unit pattern, such as Figure 2 As shown:

[0137]

[0138] Where EF is the array element pattern; N r The number of concentric ring arrays; N m is the number of array elements on the nth circle and is an integer multiple of 4 so as to divide the array into four rotationally symmetric regions; w n is the vector weight value of the nth circle array element; k is the wave number, and k = 2π / λ; r n is the radius of the nth circle; is the sampling angle in S1 space, and is the beam pointing angle;

[0139] beta(n,m) is the rotation angle of the mth unit on the nth circle relative to the x-axis. The angle difference between each unit is equal, and there is Where c is a constant, For variables, c and Determines the additional angular rotation of the unit on each circle;

[0140] For the array layout design, no amplitude and phase weighting is applied to avoid loss of transmit EIRP or receive G / T. Therefore, for all w n =1, so we have

[0141]

[0142] More preferably, in an exemplary embodiment, the targeting of the requirements of indicators including antenna sidelobe suppression, beam width, and antenna unit spacing includes:

[0143] According to the theoretical analysis results and design experience of the circular array, the number of array elements on each circle is given first, and then the radius r is determined while satisfying the beam width and being greater than the minimum spacing. n and φ n Optimize; Since the index requires different requirements for the sidelobe level when the beam is scanned to different angles, this problem is a multi-objective optimization problem and is expressed as follows:

[0144]

[0145] St.HPBW s ≥3.5°

[0146] Δd≥Δd s =9.5mm

[0147] in, is the beam pointing angle The maximum sidelobe level required when the interference avoidance mode and non-interference avoidance mode There is a difference; Peak sidelobe level calculated for an individual pattern; HPBW s is the required minimum beam width; Δd is the spacing between adjacent array elements; Δd s is the minimum required spacing between adjacent elements.

[0148] As mentioned above, the sparse array design of the concentric ring array is a multi-objective optimization problem, and a non-dominated sorting multi-objective differential evolution algorithm is proposed to optimize and solve the above problem. More preferably, in an exemplary embodiment, the optimization of the parameterized parameters by the non-dominated sorting multi-objective differential evolution algorithm includes, for example Figure 3 As shown:

[0149] a. Parameter settings: population size, maximum execution number, crossover constant, number of variables, and the parameters that affect the array pattern include the number of turns N r 、Number of turns per circle m , radius of each circle, starting angle of each circle φ n and the excitation amplitude phase value;

[0150] b. Randomly generate the initial population according to the following expression:

[0151]

[0152] c. Randomly select different individuals in the population and repeat the DE / rand / 1 mutation operation

[0153]

[0154] And get the population pop m; Scaling factor F i,j It is not a constant. For each gene point variation, it is a random number between [0,1].

[0155] d. Population m Perform a two-point crossover operation and obtain the population pop c ;

[0156] e. Evaluate population pop c The experimental individuals and the target individual x i The fitness value of

[0157] f. Pop the population m and pop c Merge and perform non-dominated sorting;

[0158] g. Calculate the congestion and sort in descending order;

[0159] h. Select the first PS individuals as the next generation population, and re-perform non-dominated sorting, crowding calculation and descending order;

[0160] Repeat steps c to h above until the required number of generations is reached, and then save the results.

[0161] Specifically, in this exemplary embodiment, the target of the array optimization is some or all of the parameters therein, an array consisting of a set of target parameters is an individual, and a collection of many individuals is a population. The formation of an individual in the initial population is random, and each target parameter will set an optimization range, and a parameter value is randomly generated within this range to form an initial individual. And repeat this step to form the entire population.

[0162] More preferably, in an exemplary embodiment, the calculation of the fitness value includes:

[0163] Calculate the array point corresponding to the parameter and the minimum distance D0 between the units;

[0164] Calculate the array pattern and main lobe half-power beam width W0;

[0165] Normalize the directional pattern, remove the main lobe part, and screen out all unqualified side lobe points with exceeding values ​​Pi and their total number Np according to the side lobe requirements;

[0166] The fitness calculation formula is as follows:

[0167]

[0168] D lim With W limIt is the minimum unit spacing and half-power bandwidth requirement. a1 and a2 are coefficients greater than 1, and b1 is a coefficient less than 1 and greater than 0. It is selected according to the situation.

[0169] Specifically, in this exemplary embodiment, in order to reduce the solution time of the planar array 3D directional pattern involved in the fitness calculation, the directional pattern of the concentric ring array is quickly solved and calculated based on the developed chirp-z transform algorithm, and the algorithm optimization time is lower.

[0170] More preferably, in an exemplary embodiment, the non-dominated sorting includes:

[0171] For each individual in the population, calculate the number of dominated individuals Ni and the number of dominated individuals Si; the domination means that all indicators of an individual A are not weaker than those of an individual B and there is an indicator that is better than that of individual B, that is, A dominates B; the indicators include bandwidth, minimum unit spacing and side lobes;

[0172] Find all individuals in the population whose Ni is 0, store them in set F1, and assign a uniform non-dominated ordinal number Ti to all individuals;

[0173] Subtract 1 from the remaining individuals S1 that are dominated by the set F1;

[0174] Repeat the above steps until all individuals are stratified into values ​​F1 to FN.

[0175] More preferably, in an exemplary embodiment, the congestion degree calculation includes:

[0176] According to the N parameters of each individual, place it in N-dimensional space;

[0177] Calculate the crowding operator Ri, where Ri is the distance between an individual and its nearest individual.

[0178] More preferably, in an exemplary embodiment, the descending order arrangement includes:

[0179] All individuals are given a non-dominated ordinal number and a crowding operator. When two individuals have different non-dominated rankings, the one with the smaller ranking number is taken. If the non-dominated rankings are the same, the less crowded individual, i.e. the one with a larger Ri, is taken, and the order is ranked in this order.

[0180] More preferably, in an exemplary embodiment, randomly rotating the antenna units in the four rotationally symmetric sub-arrays and simulating to confirm that the antenna array pattern meets expectations includes:

[0181] Obtain the coordinate position of the array element in the optimized 1 / 4 sub-array;

[0182] Use electromagnetic simulation software to paste the antenna unit model on each point in turn and number them;

[0183] Generate a random number array with a length equal to the number of units in the subarray, ranging from 0° to 360°;

[0184] Rotate the antenna units along the unit center axis in sequence according to the random angles corresponding to the numbers in the array;

[0185] Then the 1 / 4 subarray is rotated 90° around the array center axis and replicated three times to obtain the model of all units;

[0186] Complete the gaps between unit models, improve the conditions of materials, boundary conditions, and feed amplitude and phase, simulate the array, and confirm whether the antenna array radiation pattern after rotating at a random angle meets expectations.

[0187] Specifically, in this exemplary embodiment, in order to avoid the parasitic radiation problem caused by the discontinuity of the array surface, the array surface is divided into Figure 4 The four identical rotationally symmetric sub-arrays are each a complete multi-layer PCB antenna array, so the length and number of the slots are greatly reduced, minimizing the impact on the antenna's reception or radiation performance.

[0188] The dividing line is a continuous broken line, with two rotationally symmetrical 1 / 4 sub-array units on both sides of the broken line. The line segment starts from the center of the array circle, connects the centers of the closest edges of the adjacent units on both sides of the boundary, and extends to the end of the array extension. Since the array dividing line has the same central symmetry as the 1 / 4 sub-array, the method of first drawing the 1 / 4 dividing line and then rotating it is used to reduce the amount of operations.

[0189] In addition, it is necessary to further explain that in order to improve the circular polarization performance of circular polarization array antennas in wide-angle scanning and broadband satellite communication applications, the traditional method is to use Figure 5 The sequential rotation technology shown (left is the whole picture, right is the partial picture). By arranging the elements in the small subarray in a certain order to reduce the axial ratio of the beam, the circular polarization performance of a large array antenna composed of several subarrays can be effectively improved. However, this method will cause the sidelobe level of the array to deteriorate, especially when a high-level sidelobe enters the visible area from the non-visible area during scanning, causing the sidelobe level to fail to meet the index requirements.

[0190] Based on the problems existing in regular rotation array formation and the irregular array formation in this scheme, a method based on random rotation array formation is proposed. Figure 6 shown.

[0191] The application of random rotation array improves the axial ratio and gain drop during beam scanning and avoids the prominent side lobes caused by energy concentration. Figure 7 and Figure 8The following are the gain patterns of the normal and large-angle scanning of the sequential rotation array and the random rotation array, respectively. Figure 7 and Figure 8 By comparison, the random rotation array direction diagram ( Figure 8 ) Figure 7 The energy in the middle circle is too concentrated, resulting in abnormally prominent side lobes. The reason why the random rotation array scheme does not have the above problem is that it breaks the periodicity introduced by the sequential rotation array, and evenly distributes the energy of the periodically introduced high-level side lobes to other side lobes, thereby suppressing the appearance of high levels. For other side lobes, the level increase is very small.

[0192] Obviously, the above embodiments are merely examples for clear explanation, and are not intended to limit the implementation methods. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from them are still within the protection scope of the invention.

Claims

1. A design method for a randomly rotated concentric ring sparsely distributed circularly polarized phased array antenna, characterized in that: The following steps are involved: Establish a sparse concentric ring array optimization model and perform parameterization; The sparse concentric circular array optimization model includes four rotationally symmetric subarrays, each subarray includes multiple antenna units, and the parameterized parameters include the radius of 1 to N circles, the number of antenna units in 1 to N circles, and the starting angle of 1 to N circles; Using the overall antenna array pattern of the sparse concentric ring array optimization model, taking the requirements of indicators including antenna sidelobe suppression, beam width, and antenna unit spacing as the target, the parameterized parameters are optimized through a non-dominated sorting multi-objective differential evolution algorithm; The antenna elements in the four rotationally symmetric subarrays were randomly rotated, and the simulation confirmed that the antenna array pattern met expectations; The boundary between two sub-arrays is obtained, and the array is divided with a broken line as the starting point at the center of the array circle, so that the broken line passes through the center of the closest edges of adjacent units on both sides of the boundary and extends to the end of the array extension to obtain a dividing line; the dividing line is rotated 90° around the center of the circle and replicated three times.

2. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 1 is characterized in that: The step of establishing a sparse concentric ring array optimization model and parameterizing it includes: Calculate the required aperture size and number of antenna units based on gain requirements; The number of antenna units is an integer multiple of 4, and these antenna units respectively form four rotationally symmetric sub-arrays; Select the 90° sector to allocate array element positions, determine the number of circles according to the number of antenna elements, and allocate the number of elements per circle according to the arc length of each circle; The array elements in each circle are evenly arranged on a 90° arc, and the angle difference between the array elements relative to the center of the array surface is calculated; Assign a starting angle to the starting element of each circle of elements, where the starting angle is greater than 0 and not greater than the angle difference between the elements; After rotating the obtained quarter sub-array 90° around the center of the array and the surface four times, the complete antenna array is obtained. At this time, the position of each array element on the array surface is uniquely determined; The antenna array is divided into N r circles, a total of 3N parameters are required to constrain the position of each element in the antenna array, which are: the radius of 1 to N circles, namely R1, R2, ... R N , the number of antenna units in 1 to N circles is N1, N2, ... N N , the starting angle of 1 to N circles is Complete the parameterization of the array element positions in the concentric ring array.

3. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 2 is characterized by: The overall antenna array pattern of the sparse concentric ring array optimization model includes: Determine the fluctuation range of the parameters; The unit radiation pattern is obtained by simulating the antenna unit; The overall radiation pattern of the synthesized antenna array is calculated by using the coordinates and radiation patterns of the antenna units in the array: Where EF is the array element pattern; N r The number of concentric ring arrays; N m is the number of array elements on the nth circle and is an integer multiple of 4 so as to divide the array into four rotationally symmetric regions; w n is the vector weight value of the nth circle array element; k is the wave number, and k = 2π / λ; r n is the radius of the nth circle; is the sampling angle in S1 space, and is the beam pointing angle; beta(n,m) is the rotation angle of the mth unit on the nth circle relative to the x-axis. The angle difference between each unit is equal, and there is Where c is a constant, For variables, c and Determines the additional angular rotation of the unit on each circle; For the array layout design, no amplitude and phase weighting is applied to avoid loss of transmit EIRP or receive G / T. Therefore, for all w n =1, so we have 4. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 3 is characterized by: The above targets include antenna sidelobe suppression, beam width, antenna unit spacing and other indicators, including: First, the number of array elements on each circle is given, and then the radius r is set while satisfying the beam width and being greater than the minimum spacing. n and Optimize; Since the index requires different requirements for the sidelobe level when the beam is scanned to different angles, this problem is a multi-objective optimization problem and is expressed as follows: St.HPBW s ≥3.5° Δd≥Δd s =9.5mm in, is the beam pointing angle The maximum sidelobe level required when the interference avoidance mode and non-interference avoidance mode There is a difference; Peak sidelobe level calculated for an individual pattern; HPBW s is the required minimum beam width; Δd is the spacing between adjacent array elements; Δd s is the minimum required spacing between adjacent array elements.

5. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 4 is characterized by: The multi-objective differential evolution algorithm for non-dominated sorting optimizes the parameterized parameters, including: a. Parameter settings: population size, maximum execution number of generations, crossover constant, number of variables; b. Randomly generate the initial population according to the following expression: c. Randomly select different individuals in the population and repeat the DE / rand / 1 mutation operation And get the population pop m ; Scaling factor F i,j It is not a constant. For each gene point variation, it is a random number between [0,1]. d. Population m Perform a two-point crossover operation and obtain the population pop c ; e. Evaluate population pop c The experimental individuals and the target individual x i The fitness value of f. Pop the population m and pop c Merge and perform non-dominated sorting; g. Calculate the congestion and sort in descending order; h. Select the first PS individuals as the next generation population, and re-perform non-dominated sorting, crowding calculation and descending order; Repeat steps c to h above until the required number of generations is reached, and then save the results.

6. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 5 is characterized by: The calculation of the fitness value includes: Calculate the array point corresponding to the parameter and the minimum distance D0 between the units; Calculate the array pattern and main lobe half-power beam width W0; Normalize the directional pattern, remove the main lobe part, and screen out all unqualified side lobe points with exceeding standard values ​​Pi and their total number Np according to the side lobe requirements; The fitness calculation formula is as follows: D lin With W lim It is the minimum unit spacing and half-power bandwidth requirement. a1 and a2 are coefficients greater than 1, and b1 is a coefficient less than 1 and greater than 0. It is selected according to the situation.

7. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 5 is characterized by: The non-dominated sorting includes: For each individual in the population, calculate the number of dominated individuals Ni and the number of dominated individuals Si; the domination means that all indicators of an individual A are not weaker than those of an individual B and there is an indicator that is better than that of individual B, that is, A dominates B; the indicators include bandwidth, minimum unit spacing and side lobes; Find all individuals in the population whose Ni is 0, store them in set F1, and assign a uniform non-dominated ordinal number Ti to all individuals; Subtract 1 from the remaining individuals S1 that are dominated by the set F1; Repeat the above steps until all individuals are stratified into values ​​F1 to FN.

8. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 7 is characterized in that: The congestion calculation includes: According to the N parameters of each individual, place it in N-dimensional space; Calculate the crowding operator Ri, where Ri is the distance between an individual and its nearest individual.

9. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 8, characterized in that: The descending order includes: All individuals are given a non-dominated ordinal number and a crowding operator. When two individuals have different non-dominated rankings, the one with the smaller ranking number is taken. If the non-dominated rankings are the same, the less crowded individual, i.e. the one with a larger Ri, is taken, and the order is ranked in this order.

10. The design method of the randomly rotated concentric ring sparse array circularly polarized phased array antenna according to claim 1, characterized in that: The antenna elements in the four rotationally symmetric sub-arrays are randomly rotated, and the simulation confirms that the antenna array pattern meets expectations, including: Obtain the coordinate position of the array element in the optimized 1 / 4 sub-array; Use electromagnetic simulation software to paste the antenna unit model on each point in turn and number them; Generate a random number array with a length equal to the number of units in the subarray, ranging from 0° to 360°; Rotate the antenna units along the unit center axis in sequence according to the random angles corresponding to the numbers in the array; Then the 1 / 4 subarray is rotated 90° around the array center axis and replicated three times to obtain the model of all units; Complete the gaps between unit models, improve the conditions of materials, boundary conditions, and feed amplitude and phase, simulate the array, and confirm whether the antenna array radiation pattern after rotating at a random angle meets expectations.

Images