Electronic bidding and tendering system optimization method

Abstract

The invention discloses an electronic bidding system optimization method, and relates to the technical field of bidding system optimization, and the method comprises the steps: representing the selection state of each bidding party as a quantum bit, dynamically adjusting the selection state of the quantum bit through quantum gate operation, enabling the initial state of the quantum bit to be a uniform superposition state, and enabling the initial state of the quantum bit to be a uniform superposition state; after the optimizer calculates an objective function value of each quantum bit, parameter values in the quantum bits are updated, current parameter values of all the quantum bits are obtained, and after secondary optimization is carried out on all the quantum bits through a genetic algorithm, the genetic algorithm outputs a plurality of groups of quantum bit sets; and after the output end combines the multiple groups of quantum bit sets to obtain the scheme set, the quantum bit selection state in the scheme set is automatically selected as a final selection scheme to be output. According to the optimization method, exploration of the understanding space is accelerated through parallel processing, compared with a traditional method, an approximate optimal solution can be found in a shorter time, and the optimization method is particularly suitable for the situation that the number of bidding parties is large.

Description

Technical Field

[0001] The present invention relates to the technical field of optimizing bidding systems, and specifically relates to a method for optimizing electronic bidding systems. Background Art

[0002] With the acceleration of the global informatization process, more and more bidding activities are carried out through electronic platforms. Especially in the fields of construction, engineering, government procurement, etc., the traditional bidding methods can no longer meet the requirements of efficiency and transparency. Therefore, electronic bidding systems have emerged and quickly become the industry standard. The optimization of electronic bidding systems is usually aimed at improving the system efficiency, reducing human operation errors, enhancing data security, and ensuring a more transparent and fair bidding process.

[0003] The existing technologies have the following defects:

[0004] 1. Traditional optimization methods (such as linear programming, integer programming, or heuristic algorithms) have extremely high computational complexity when facing multi-dimensional and multi-objective combinatorial optimization problems in bidding systems. Especially when the number of bidders increases, the solution space of the optimization problem grows exponentially, resulting in low computational efficiency and thus reducing the optimization efficiency of the bidding system;

[0005] 2. Traditional optimization methods often encounter the problem of "local optimal solutions", that is, the optimization process may fall into a certain local optimal point and fail to find the global optimal solution. Especially traditional optimization algorithms such as gradient descent are easily affected by the initial values, resulting in inaccurate results.

[0006] Based on this, the present invention proposes a method for optimizing an electronic bidding system, which accelerates the exploration of the solution space through parallel processing. Compared with traditional methods, it can find approximate optimal solutions in a shorter time, especially suitable for the case of a large number of bidders, and further reduces the influence of local optimal solutions through the diversity of the population, enhances the global search ability, and improves the accuracy of the calculation results. Summary of the Invention

[0007] The purpose of the present invention is to provide a method for optimizing an electronic bidding system to solve the deficiencies in the background art.

[0008] To achieve the above purpose, the present invention provides the following technical solution: A method for optimizing an electronic bidding system, the optimization method includes the following steps:

[0009] The processing end obtains all the bidder information related to the bidding project, represents the selection status of each bidder as a qubit, and dynamically adjusts the selection status of the qubit through quantum gate operations to make the initial state of the qubit a uniform superposition state;

[0010] After the optimizer calculates the objective function value of each qubit, it updates the parameter values in the qubit;

[0011] When the optimizer converges, it obtains the current parameter values of all qubits, and after performing secondary optimization on all qubits through a genetic algorithm, the genetic algorithm outputs several sets of qubit sets. After the output end combines the multiple sets of qubit sets to obtain a solution set, it automatically selects the qubit selection state in the solution set as the final selection solution for output.

[0012] In a preferred embodiment, the selection state of the qubit is dynamically adjusted through quantum gate operations to make the initial state of the qubit a uniform superposition state, including the following steps:

[0013] Suppose there are N bidders, corresponding to N qubits, and the initial selection state of each qubit is ∣0>, indicating that the bidder is not selected;

[0014] Apply the Hadamard gate to each qubit to change the qubit from the ∣0> state to a uniform superposition state. The Hadamard gate operation on the qubit is expressed as: H|0> is the initial selection state of the qubit;

[0015] For the initial selection states of N qubits, after the Hadamard gate operation, all qubits are in a uniform superposition state, representing the superposition of the selection states of all bidders: In the formula, |ψ(0)> represents the initial selection state of the qubit, is the normalization coefficient of the uniform superposition state of the qubit, represents the superposition of the selection states of all qubits, and i ranges from 0 to 2 N -1, representing all bidder selection combinations.

[0016] In a preferred embodiment, after the optimizer calculates the objective function value of each qubit, it updates the parameter values in the qubit, including the following steps:

[0017] The optimizer obtains the selection cost, quality penalty index, and delivery index of each qubit, and calculates the objective function value of each qubit;

[0018] Change the selection state of the first bidder in the qubit, then calculate the objective function value of the qubit after the change, and compare the objective function value of the qubit after the change with the objective function value of the qubit before the change;

[0019] If the objective function value of the qubit after the change is less than that of the qubit before the change, then retain the qubit with the changed selected state. If the objective function value of the qubit after the change is greater than or equal to that of the qubit before the change, then retain the qubit with the selected state before the change;

[0020] After performing the selected state change and retention analysis for all bidders in the qubit in sequence, the final selected state of the qubit is obtained, that is, the qubit parameter value.

[0021] In a preferred embodiment, calculate the objective function value of each qubit, and the expression is: In the formula, f is the objective function value, x i is the selected state of the i-th bidder, and x i ∈ {0, 1}, T_C is the selection cost, Q_P is the quality penalty index, D_P is the delivery index, α, β, and γ are the proportionality coefficients of the selection cost, quality penalty index, and delivery index respectively, and α, β, and γ are all greater than 0.

[0022] In a preferred embodiment, the acquisition logic of the delivery index is as follows: Obtain the supply quantities of all bidders with the selected state of ∣1> in the qubit, sum up the supply quantities of all bidders with the selected state of ∣1> to obtain the total supply quantity, and obtain the supply quantity difference by subtracting the required supply quantity of the construction project from the total supply quantity;

[0023] Obtain the expected supply times of all bidders with the selected state of ∣1>, and obtain the required supply time of the construction project. If the total supply quantity of all bidders with the selected state of ∣1> is greater than or equal to the required supply quantity of the construction project, then use the maximum expected supply time among all bidders with the selected state of ∣1> as the latest expected supply time. If the total supply quantity of all bidders with the selected state of ∣1> is less than the required supply quantity of the construction project, then use twice the maximum expected supply time among all bidders with the selected state of ∣1> as the latest expected supply time, and obtain the remaining duration by subtracting the latest expected supply time from the required supply time;

[0024] Perform normalization processing on the supply quantity difference and the remaining duration, map the value ranges of the supply quantity difference and the remaining duration to between [0, 1], and sum up the normalized supply quantity difference and the remaining duration to obtain the delivery index.

[0025] In a preferred embodiment, the acquisition logic of the selection cost is as follows: Obtain the cost values of all bidders with the selected state of ∣1> in the qubit, and sum up the cost values of all bidders with the selected state of ∣1> to obtain the selection cost;

[0026] The calculation logic of the quality penalty index is as follows: Obtain the historical return rates of all bidders with the selected state ∣1> in the qubits, and sum up the historical return rates of all bidders with the selected state ∣1> to obtain the quality penalty index.

[0027] In a preferred embodiment, after performing secondary optimization on all qubits through a genetic algorithm, the genetic algorithm outputs several sets of qubit sets, including the following steps:

[0028] When the number of optimization times of the optimizer is equal to the optimization times threshold or the objective function value of any qubit is less than or equal to the second threshold, it is determined that the optimizer converges, and the current parameter values of all qubits are obtained;

[0029] After obtaining all qubits and their corresponding objective function values, perform a retention operation on all qubits, and perform an information exchange operation on the retained qubits and then a mutation operation;

[0030] Establish a qubit set for all qubits after the mutation operation is completed, and repeat the deletion, information exchange operation, and mutation operation for all qubits in the qubit set;

[0031] When the objective function value of any qubit is less than or equal to the first threshold or the number of loops is equal to the loop number threshold, output several sets of qubit sets.

[0032] In a preferred embodiment, the processing logic of the retention operation is as follows: Obtain all qubits and their corresponding objective function values, sum up the objective function values of all qubits to obtain the denominator value, divide the objective function value by the denominator value to obtain the deletion probability of the qubit, map the deletion probabilities of all qubits to the sector areas of a virtual roulette, start the rotation of the virtual roulette, and when the rotation stops, delete the qubit corresponding to the sector area pointed to by the virtual pointer from the virtual roulette. Repeat the above steps, and when the number of qubits remaining on the virtual roulette is equal to the number threshold, retain the qubits remaining on the virtual roulette.

[0033] In a preferred embodiment, after the output end combines multiple sets of qubit sets to obtain a solution set, it automatically selects the qubit selection state in the solution set as the final selection solution for output, including the following steps:

[0034] The output end combines multiple sets of qubit sets to obtain a solution set, and sequentially compares the objective function values of all qubits in the solution set, and automatically selects the qubit selection state with the smallest objective function value as the final selection solution for output.

[0035] In a preferred embodiment, the processing end obtains all bidder information related to the tender project, and represents the selection state of each bidder as a qubit, including the following steps:

[0036] The processing end obtains the relevant information of all bidders from the bidding system. The relevant information includes the identifiers of the bidders, the bids or costs of the bidders, the quality scores of the products or services provided by the bidders, the delivery times of the bidders, the evaluation of the performance capabilities of the bidders, and whether the bidders meet the technical requirements of the project.

[0037] The relevant information of the bidders is used as input data. A qubit is assigned to each bidder to represent whether to select the bidder. Given N bidders, there will be N qubits. The selection state of each qubit is represented as: ∣0>: indicating that the bidder is not selected; ∣1>: indicating that the bidder is selected.

[0038] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0039] 1. In the present invention, the selection state of each bidder is represented as a qubit, and the selection state of the qubit is dynamically adjusted through quantum gate operations, so that the initial state of the qubit is a uniform superposition state. After the optimizer calculates the objective function value of each qubit, the parameter value in the qubit is updated. When the optimizer converges, the current parameter values of all qubits are obtained, and all qubits are secondarily optimized through a genetic algorithm. After the genetic algorithm outputs several sets of qubit sets, the output end combines the multiple sets of qubit sets to obtain a solution set, and automatically selects the qubit selection state in the solution set as the final selection solution for output. The optimization method accelerates the exploration of the solution space through parallel processing. Compared with traditional methods, it can find an approximate optimal solution in a shorter time, especially suitable for the case of a large number of bidders, and further reduces the influence of local optimal solutions through the diversity of the population, enhances the global search ability, and improves the accuracy of the calculation results.

[0040] 2. In the present invention, by applying the Hadamard operation to each qubit, we initialize the selection state of each bidder to a uniform superposition state, so that the state of the entire system represents the superposition of all possible bidder selection solutions. This method enables all selection states to be considered simultaneously through quantum parallelism, thus providing a rich search space for subsequent optimization operations (such as quantum gate operations, objective function optimization, etc.) in the quantum optimization process.

[0041] 3. In the present invention, the genetic algorithm is used to secondarily optimize the selection state of the qubits. By using deletion, information exchange, and mutation operations, the optimal solution can be found from multiple selection solutions. This method effectively combines the advantages of quantum computing and classical optimization algorithms, can handle complex optimization problems, such as the bidder selection problem, and finally obtains the optimal solution that meets the requirements of cost, delivery period, and quality. Description of the Drawings

[0042] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments described in the present invention. For those of ordinary skill in the art, other accompanying drawings can also be obtained based on these drawings.

[0043] Figure 1 It is a flowchart of the method of the present invention. Specific embodiments

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0045] Embodiment 1: Please refer to Figure 1 As shown, the optimization method of the electronic bidding and tendering system in this embodiment includes the following steps:

[0046] The processing end obtains all the information of the bidders related to the bidding project, represents the selection status of each bidder as a qubit, dynamically adjusts the selection status of the qubit through quantum gate operations, so that the initial state of the qubit is a uniform superposition state. After the optimizer calculates the objective function value of each qubit, it updates the parameter value in the qubit. When the optimizer converges, it obtains the current parameter values of all qubits, and after performing secondary optimization on all qubits through the genetic algorithm, the genetic algorithm outputs several sets of qubit sets. After the output end combines the multiple sets of qubit sets to obtain a solution set, it automatically selects the qubit selection status in the solution set as the final selection solution for output.

[0047] In this application, the selection status of each bidder is represented as a qubit. By performing quantum gate operations, the selection status of the qubit is dynamically adjusted to make the initial state of the qubit a uniform superposition state. After the optimizer calculates the objective function value of each qubit, the parameter value in the qubit is updated. When the optimizer converges, the current parameter values of all qubits are obtained, and all qubits are secondarily optimized by a genetic algorithm. After the genetic algorithm outputs several sets of qubit sets, the output end combines the multiple sets of qubit sets to obtain a solution set, and automatically selects the qubit selection status in the solution set as the final selection solution for output. The optimization method accelerates the exploration of the solution space through parallel processing. Compared with traditional methods, it can find an approximate optimal solution in a shorter time, especially suitable for the case of a large number of bidders, and further reduces the influence of local optimal solutions through the diversity of the population, enhances the global search ability, and improves the accuracy of the calculation results.

[0048] The optimization system includes an adjustment module, an update module, and a solution output module:

[0049] Adjustment module: Obtain all bidder information related to the tender project, represent the selection status of each bidder as a qubit, dynamically adjust the selection status of the qubit through quantum gate operations to make the initial state of the qubit a uniform superposition state, and send the qubit information to the update module;

[0050] Update module: After the optimizer calculates the objective function value of each qubit, update the parameter value in the qubit, and send the optimization result of the optimizer to the solution output module;

[0051] Solution output module: When the optimizer converges, obtain the current parameter values of all qubits, secondarily optimize all qubits by a genetic algorithm, after the genetic algorithm outputs several sets of qubit sets, combine the multiple sets of qubit sets to obtain a solution set, and automatically select the qubit selection status in the solution set as the final selection solution for output.

[0052] Embodiment 2: The processing end obtains all bidder information related to the tender project and represents the selection status of each bidder as a qubit, including the following steps:

[0053] The processing end first needs to obtain all relevant information of all bidders from the tender system, including but not limited to:

[0054] Bidder name: The identifier of each bidder.

[0055] Cost: The quotation or cost of each bidder.

[0056] Quality: The quality score of the products or services provided by each bidder.

[0057] Delivery period: The delivery time of each bidder.

[0058] Performance ability: The performance ability evaluation of each bidder, which may be a score based on historical data.

[0059] Technical requirements and other constraints: Whether each bidder meets the technical requirements or other specific constraints of the project.

[0060] This data will be used as input data for the subsequent qubit representation and optimization process. Each bidder is assigned a qubit to represent whether the bidder is selected.

[0061] Suppose we have N bidders, then we will have N qubits. The selection state of each qubit can be represented as:

[0062] ∣0>: Represents not selecting this bidder.

[0063] ∣1>: Represents selecting this bidder.

[0064] Suppose a project has 4 bidders, and the specific information is shown in Table 1:

[0065] Table 1 Information Table of Bidders

[0066]

[0067] Assign a qubit q1, q2, q3, q4 to each bidder in Table 1, where:

[0068] q1 represents the selection state of bidder A, q2 represents the selection state of bidder B, q3 represents the selection state of bidder C, and q4 represents the selection state of bidder D;

[0069] The selection state of each qubit can be represented as:

[0070] ∣0>: Represents not selecting this bidder.

[0071] ∣1>: Represents selecting this bidder.

[0072] Therefore, if we select bidders A and D and do not select B and C, then the state of the qubits will be: |q1,q2,q3,q4> = |1,0,0,1>, which means that bidders A and D are selected while B and C are not selected.

[0073] Dynamically adjust the selection state of the qubits through quantum gate operations to make the initial state of the qubits a uniform superposition state, including the following steps:

[0074] Suppose there are N bidders, then there will be N qubits, and the initial state of each qubit is ∣0>, indicating that the bidder is not selected. To make the state of the qubits represent all possible combinations of bidders (i.e., selecting or not selecting a bidder), we apply the Hadamard gate (H gate) to each qubit, changing its state from ∣0> to a uniform superposition state. The Hadamard gate operation on a single qubit is expressed as: Understanding the selection state of the qubit after applying the Hadamard gate means that the state of the qubit becomes a uniform superposition between ∣0> and ∣1>.

[0075] For the initial selection state of N qubits, after the Hadamard gate operation, all qubits are in a uniform superposition state, representing the superposition of all possible bidder selection states: In the formula, |ψ(0)> represents the initial selection state of the qubit, that is, the quantum state after the Hadamard gate operation. is the normalization coefficient of the uniform superposition state of the qubit. Normalization ensures that the sum of the probability amplitudes of the quantum state is 1, so that the probability is correct during measurement. represents the superposition of all possible qubit selection states, where i ranges from 0 to 2 N - 1, representing all possible combinations of bidder selections, and each |i> is a quantum state representing a different bidder selection.

[0076] If we have multiple bidders, each bidder is assigned a qubit, and the Hadamard operation is applied to the initial state of each qubit, then the state of the entire system will be the product of the states of these individual qubits. Initialize each qubit to ∣0> (indicating that no bidder is selected initially), and apply the Hadamard operation to each qubit. For the qubit of the first bidder, the operation is as follows: Similarly, apply the Hadamard operation to the qubit of the second bidder:

[0077] Since the operations of the qubits are independent, we combine the states of the two qubits into the state of a single system. The state of the overall system is:

[0078] This state can be expanded as: In this two - qubit system, after the Hadamard operation, the state of the system is a uniform superposition of the choice states of each bidder. Specifically, the system state is a superposition of all 4 possible choice states (|00>, |01>, |10>, |11>), and the probability amplitudes of each state are equal. Each bidder has a 50% probability of choosing, where ∣0> represents (not chosen) and ∣1> represents (chosen).

[0079] If we measure the system, the possible final results are:

[0080] ∣00>: Neither bidder is chosen. ∣01>: Bidder 1 is chosen and bidder 2 is not chosen. ∣10>: Bidder 1 is not chosen and bidder 2 is chosen. ∣11>: Both bidders are chosen. The probability of each result is 25% (since each qubit has a 50% probability of being in the ∣0> or ∣1> state).

[0081] In this application, by applying the Hadamard operation to each qubit, we initialize the choice state of each bidder to a uniform superposition state, such that the state of the entire system represents a superposition of all possible bidder choice scenarios. This method, through quantum parallelism, enables all choice states to be considered simultaneously, thus providing a rich search space for subsequent optimization operations (such as quantum gate operations, objective function optimization, etc.) in the quantum optimization process.

[0082] After the optimizer calculates the objective function value of each qubit, it updates the parameter values in the qubit, including the following steps:

[0083] The optimizer obtains the selection cost, quality penalty index, and delivery index of each qubit, and calculates the objective function value of each qubit. The expression is:

[0084] In the formula, f is the objective function value, x i is the choice state of the i - th bidder, and x i ∈{0,1}, T_C is the selection cost, that is, the total cost of all selected bidders, Q_P is the quality penalty index, D_P is the delivery index, α, β, γ are the proportionality coefficients of the selection cost, quality penalty index, and delivery index respectively, and α, β, γ are all greater than 0.

[0085] The acquisition logic of the selection cost is: Obtain the cost values of all bidders with the choice state ∣1> in the qubit, and sum up the cost values of all bidders with the choice state ∣1> to obtain the selection cost. The smaller the selection cost, the lower the cost of the selected bidders, and the more favorable it is. In the optimization process, the goal is to minimize the selection cost as much as possible.

[0086] The logic for obtaining the delivery index is as follows: obtain the supply of all bidders whose selection state is |1> in the quantum bit, sum up the supply of all bidders whose selection state is |1> to obtain the total supply, and obtain the supply difference by subtracting the demand supply of the construction project from the total supply. The smaller the supply difference, the less the total supply meets the demand supply of the construction project under the current selection state of the quantum bit, obtain the estimated delivery time of all bidders whose selection state is |1>, and obtain the demand delivery time of the construction project. If the total supply of all bidders whose selection state is |1> is greater than or equal to the demand supply of the construction project, then the maximum estimated delivery time among all bidders whose selection state is |1> is calculated. As the latest estimated delivery time, if the total supply of all bidders with a selection state of |1> is less than the demand supply of the construction project, then the twice the maximum estimated delivery time of all bidders with a selection state of |1> is taken as the latest estimated delivery time (indicating that the bidder needs to supply for two cycles), and the surplus time is obtained by subtracting the latest estimated delivery time from the demand delivery time. The smaller the surplus time, the tighter the delivery time is under the current selection state of the quantum bit. The supply difference and the surplus time are normalized so that the value range of the supply difference and the surplus time is mapped to [0,1]. The normalized supply difference and surplus time are summed to obtain the delivery index. The delivery index is calculated based on the supply difference and the surplus time. The smaller the supply difference and the surplus time, the tighter the supply plan is. The smaller the delivery index value, the tighter the delivery time is, and the less it meets the needs of the construction project. Therefore, the delivery index needs to be maximized, which means that under the current selection state, the less tight the supply plan is, the more it can meet the needs. In other words, the larger the delivery index is, the more the supply capacity of the selected bidder matches the needs, and the more sufficient the delivery time is.

[0087] To explain the above process in detail, we will use a specific example to illustrate how to calculate the total supply, supply difference, latest estimated supply time and surplus time under the quantum bit selection state.

[0088] Assumptions:

[0089] Demand supply for construction projects: 5,000 units;

[0090] Demand and delivery time for construction projects: 60 days;

[0091] Bidder information:

[0092] Assume there are 4 bidders, the specific information is as follows:

[0093] Table 1 Bidder information

[0094] Bidder Selection Status Supply Quantity (unit) Estimated Supply Time (days) A 1 2000 30 B 0 1000 40 C 1 1500 50 D 1 1200 60

[0095] The selected bidders with a status of 1 are A, C, and D (indicating that these bidders are selected). The other bidders (B) are not selected, and the supply quantity is 0.

[0096] First, we obtain the supply quantities of all bidders with a selection status of ∣1> (i.e., the selected bidders). These bidders are A, C, and D, and their supply quantities are 2000, 1500, and 1200 units respectively. Therefore, the total supply quantity is: Total supply quantity = 2000 + 1500 + 1200 = 4700 units. The supply quantity difference is the total supply quantity minus the required supply quantity of the construction project: Supply quantity difference = Total supply quantity - Required supply quantity = 4700 - 5000 = -300 units. This result indicates that the current selection plan is short of 300 units in terms of supply quantity, that is, the supply quantity is still lacking 300 units to meet the demand.

[0097] Next, we obtain the estimated supply times of all bidders with a selection status of ∣1>. The bidders with a selection status of ∣1> are A, C, and D, and their estimated supply times are 30 days, 50 days, and 60 days respectively. Now, we determine how to calculate the latest estimated supply time based on the supply quantity difference:

[0098] If the total supply quantity is greater than or equal to the required supply quantity, the latest estimated supply time is the maximum estimated supply time among the selected bidders. If the total supply quantity is less than the required supply quantity, the latest estimated supply time is twice the maximum estimated supply time among the selected bidders.

[0099] In this example, the total supply quantity is less than the required supply quantity (4700 units < 5000 units), so the "twice the maximum estimated supply time" rule needs to be used. The maximum estimated supply time among the bidders with a selection status of ∣1> is 60 days (bidder D). Therefore, the latest estimated supply time is: 120 days.

[0100] The remaining time is the required supply time of the construction project minus the latest estimated supply time. Remaining time = Required supply time - Latest estimated supply time = 60 - 120 = -60 days.

[0101] Result interpretation: The supply quantity difference is -300 units, indicating that the supply quantity of the currently selected bidders is still 300 units short of meeting the requirements of the construction project. The latest estimated supply time is 120 days, which means that if the selected bidders still maintain the current supply quantity, the supply time of the entire project will be delayed to 120 days, which is 60 days more than the required supply time of the project (60 days). The remaining time is -60 days, indicating that under the current selection status of the qubits, the supply time is already 60 days tight compared to the required supply time of the construction project, that is, the supply time is too tight.

[0102] In this example, by calculating the supply quantity, supply quantity difference, latest expected supply time, and remaining duration of the bidders with the selected state of ∣1>, we can quantify whether the current plan can meet the requirements of the construction project and adjust the selection of bidders according to the optimized selected state, so as to optimize the supply time and the degree of meeting the requirements.

[0103] The calculation logic of the quality penalty index is as follows: Obtain the historical return rates of all bidders with the selected state of ∣1> in the qubits, sum up the historical return rates of all bidders with the selected state of ∣1> to obtain the quality penalty index, and the quality penalty index is the sum of the historical return rates of all bidders with the selected state of ∣1>. The lower the historical return rate, the better the quality of the bidder, and the smaller the quality penalty index, the better the quality and the more compliant the performance of the bidder.

[0104] According to the calculation expression of the objective function value, the smaller the objective function value, the better the overall performance of the qubits;

[0105] Change the selected state of the first bidder in the qubits (for example, change the selected state of the bidder from ∣0> not selected to ∣1> selected), then calculate and obtain the objective function value of the qubits after the change, and compare the objective function value of the qubits after the change with the objective function value of the qubits before the change. If the objective function value of the qubits after the change is less than the objective function value of the qubits before the change, then retain the qubits with the changed selected state. If the objective function value of the qubits after the change is greater than or equal to the objective function value of the qubits before the change, then retain the qubits with the selected state before the change. After performing the selected state change and retention analysis on all bidders in the qubits in turn, the final selected state of the qubits, that is, the qubit parameter value, is obtained.

[0106] The selected state of the qubits will evolve according to the updated parameters. After a series of quantum gate operations, the state of the qubits will change, thus optimizing the state of the entire system. Each updated qubit state will result in a new bidder selection plan. During the quantum measurement process, the system will output a definite selection plan according to the state of the qubits.

[0107] When the optimizer converges, obtain the current parameter values of all qubits, and after performing secondary optimization on all qubits through the genetic algorithm, the genetic algorithm outputs several sets of qubit sets, including the following steps:

[0108] When the number of optimization times of the optimizer is equal to the optimization times threshold or there is a qubit with an objective function value less than or equal to the second threshold, it is determined that the optimizer converges, and the current parameter values of all qubits are obtained.

[0109] Obtain all qubits and their corresponding objective function values. Sum the objective function values of all qubits to obtain the denominator value. Divide the objective function value by the denominator value to obtain the deletion probability of the qubit. Map the deletion probabilities of all qubits to the sector areas of the virtual roulette. The larger the deletion probability, the larger the corresponding sector area. Start the rotation of the virtual roulette. When the rotation stops, delete the qubit corresponding to the sector area pointed to by the virtual pointer from the virtual roulette. Repeat the above steps. When the number of qubits remaining on the virtual roulette is equal to the number threshold, retain the qubits remaining on the virtual roulette, perform an information exchange operation on the retained qubits, and then perform a mutation operation. Establish a qubit set for all qubits after the mutation operations are completed. Repeat the deletion, information exchange operation, and mutation operation for all qubits in the qubit set. When the objective function value of any qubit is less than or equal to the first threshold or the number of loops is equal to the loop number threshold, output several groups of qubit sets.

[0110] Suppose we have 3 qubits, and the selected states of each qubit represent different combinations of bidders' selections. The initial selected states of the qubits are as follows:

[0111] Selected state of qubit 1: ∣1,0,1,1,0> (select bidders A, C, and D, do not select B and E);

[0112] Selected state of qubit 2: ∣0,1,1,0,1> (select bidders B, C, and E, do not select A and D);

[0113] Selected state of qubit 3: ∣1,1,0,0,1> (select bidders A, B, and E, do not select C and D).

[0114] The information exchange operation simulates the crossover operation in the genetic algorithm and generates new offspring qubits by exchanging the selected states of two qubits. We randomly select some genes (i.e., selected states) from two parent qubits for exchange to generate new qubits.

[0115] Suppose we select some selected states from qubit 1 and qubit 2 for exchange. The specific operation is as follows:

[0116] The selected state of qubit 1 is: ∣1,0,1,1,0>; the selected state of qubit 2 is: ∣0,1,1,0,1>; we decide to exchange the selected states of the first 3 bits. After the exchange, the new selected state of the qubit is:

[0117] New selected state of qubit 1: ∣1,1,1,0,1> (the first 3 bits come from qubit 2, and the last 2 bits come from qubit 1);

[0118] The selected state of the new qubit 2: ∣0,0,1,1,0> (the first 3 bits are from qubit 1, and the last 2 bits are from qubit 2);

[0119] The set of qubits after information exchange is:

[0120] The new qubit 1: ∣1,1,1,0,1>; the new qubit 2: ∣0,0,1,1,0>.

[0121] The mutation operation simulates the mutation in the genetic algorithm, that is, randomly changing some selected states of a qubit. This operation is to increase the diversity of the solution space and prevent falling into local optimal solutions.

[0122] Suppose we perform mutation operations on qubit 1 and qubit 2:

[0123] The selected state of qubit 1 is: ∣1,1,1,0,1>; the selected state of qubit 2 is: ∣0,0,1,1,0>.

[0124] Mutation of qubit 1: We randomly select the 4th bit for mutation and change it from 0 to 1. The selected state of the mutated qubit 1 is: ∣1,1,1,1,1>;

[0125] Mutation of qubit 2: We randomly select the 2nd bit for mutation and change it from 0 to 1. The selected state of the mutated qubit 2 is: ∣0,1,1,1,0>.

[0126] According to the selected states of the new qubits, we can recalculate the objective function (such as selection cost, delivery date, quality score, etc.) and evaluate the advantages and disadvantages of each qubit. Then, we select the optimal qubit state according to the objective function value as the basis for the next optimization.

[0127] This application performs secondary optimization on the selected states of qubits through the genetic algorithm. By using deletion, information exchange, and mutation operations, the optimal solution can be found from multiple selection schemes. This method effectively combines the advantages of quantum computing and classical optimization algorithms and can handle complex optimization problems, such as the selection problem of bidders, and finally obtains the optimal solution that meets the cost, delivery date, and quality requirements.

[0128] After the output end combines multiple sets of qubit set combination acquisition schemes to obtain a set of schemes, it automatically selects the selected state of the qubits in the set of schemes as the final selection scheme for output, including the following steps:

[0129] The output end combines multiple sets of qubit set combination acquisition schemes and sequentially compares the objective function values of all qubits in the set of schemes, and automatically selects the selected state of the qubit with the smallest objective function value as the final selection scheme for output.

[0130] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulation to get a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0131] In the description of this specification, the description referring to terms such as "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in connection with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0132] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to only the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principle and practical application of the present invention, so that those skilled in the relevant technical field can well understand and utilize the present invention.

Claims

1. An electronic bidding system optimization method, characterized in that: The optimization method comprises the following steps: The processing end obtains all bidder information related to the bidding project, and represents the selection state of each bidder as a quantum bit. The selection state of the quantum bit is dynamically adjusted through quantum gate operations, so that the initial state of the quantum bit is a uniform superposition state. After the optimizer calculates the objective function value for each qubit, it updates the parameter value in the qubit; When the optimizer converges, the current parameter values ​​of all quantum bits are obtained, and all quantum bits are optimized twice through the genetic algorithm. The genetic algorithm outputs several groups of quantum bit sets. After the output end combines multiple groups of quantum bit sets to obtain a solution set, it automatically selects the quantum bit selection state in the solution set as the final selection solution output.

2. The electronic bidding system optimization method according to claim 1 is characterized in that: Dynamically adjusting the selection state of the quantum bit through quantum gate operation so that the initial state of the quantum bit is a uniform superposition state includes the following steps: Suppose there are N bidders, corresponding to N qubits, and the initial selection state of each qubit is |0>, indicating that the bidder is not selected; Apply a Hadamard gate to each qubit to change the qubit from the |0> state to a uniform superposition state. The Hadamard gate operation on the qubit is expressed as: H|0> is the initial selection state of the quantum bit; For the initial selection state of N qubits, after the Hadamard gate operation, all qubits are in a uniform superposition state, indicating the superposition of the selection states of all bidders: Where |ψ(0)> represents the initial selection state of the quantum bit, is the normalization coefficient of the uniform superposition state of the quantum bit, represents the superposition of all qubit selection states, i from 0 to 2 N -1, indicating that all bidders select the combination.

3. The electronic bidding system optimization method according to claim 2 is characterized in that: After the optimizer calculates the objective function value for each qubit, it updates the parameter value in the qubit, including the following steps: The optimizer obtains the selection cost, quality penalty index, and delivery index of each qubit and calculates the objective function value of each qubit; The selection state of the first bidder in the quantum bit is changed, and then the objective function value of the quantum bit after the change is calculated and obtained, and the objective function value of the quantum bit after the change is compared with the objective function value of the quantum bit before the change; If the objective function value of the qubit after the change is less than the objective function value of the qubit before the change, the qubit after the change is selected is retained; if the objective function value of the qubit after the change is greater than or equal to the objective function value of the qubit before the change, the qubit before the change is selected is retained; After performing selection state change and retention analysis on all bidders in the quantum bit in turn, the final quantum bit selection state, that is, the quantum bit parameter value, is obtained.

4. The electronic bidding system optimization method according to claim 3 is characterized in that: Calculate the objective function value of each quantum bit, the expression is: In the formula, f is the objective function value, x i is the choice state of the ith bidder, and x i ∈{0,1}, T_C is the selection cost, Q_P is the quality penalty index, D_P is the delivery index, α, β, γ are the proportional coefficients of the selection cost, quality penalty index and delivery index respectively, and α, β, γ are all greater than 0.

5. The electronic bidding system optimization method according to claim 4 is characterized in that: The logic for obtaining the delivery index is as follows: obtaining the supply quantities of all bidders whose selection states are |1> in the quantum bits, summing up the supply quantities of all bidders whose selection states are |1> to obtain the total supply quantity, and obtaining the supply quantity difference by subtracting the demand supply quantity of the construction project from the total supply quantity; Obtain the estimated delivery time of all bidders whose selection status is |1>, and obtain the demand delivery time of the construction project. If the total supply of all bidders whose selection status is |1> is greater than or equal to the demand supply of the construction project, then the maximum estimated delivery time among all bidders whose selection status is |1> is taken as the latest estimated delivery time. If the total supply of all bidders whose selection status is |1> is less than the demand supply of the construction project, then the twice maximum estimated delivery time among all bidders whose selection status is |1> is taken as the latest estimated delivery time. Obtain the remaining time by subtracting the latest estimated delivery time from the demand delivery time. The supply quantity difference and the surplus time are normalized so that their value ranges are mapped to [0,1]. The normalized supply quantity difference and surplus time are summed to obtain the delivery index.

6. The electronic bidding system optimization method according to claim 5 is characterized in that: The logic for obtaining the selection cost is as follows: obtaining the cost values ​​of all the qubits whose selection states are |1> the bidder, and summing up the cost values ​​of all the qubits whose selection states are |1> the bidder to obtain the selection cost; The calculation logic of the quality penalty index is: obtain the historical return rate of all the selection states of |1> the bidder in the quantum bits, and sum up the historical return rates of all the selection states of |1> the bidder to obtain the quality penalty index.

7. The electronic bidding system optimization method according to claim 6 is characterized in that: After all qubits are optimized twice by the genetic algorithm, the genetic algorithm outputs several groups of qubit sets, including the following steps: When the optimizer optimization times are equal to the optimization times threshold or the objective function value of a quantum bit is less than or equal to the second threshold, the optimizer is judged to be converged, and the current parameter values ​​of all quantum bits are obtained; After obtaining all quantum bits and the corresponding objective function values, all quantum bits are retained, and the retained quantum bits are subjected to information exchange and then mutation operations; A quantum bit set is established for all quantum bits that have undergone mutation operations, and deletion, information exchange operations, and mutation operations are repeatedly performed on all quantum bits in the quantum bit set; When there is any quantum bit whose objective function value is less than or equal to the first threshold or the number of cycles is equal to the cycle number threshold, a plurality of quantum bit sets are output.

8. The electronic bidding system optimization method according to claim 7 is characterized in that: The processing logic of the retention operation is: obtain all quantum bits and the corresponding objective function values, sum the objective function values ​​of all quantum bits to obtain the denominator value, divide the objective function value by the denominator value to obtain the deletion probability of the quantum bit, map the deletion probability of all quantum bits to the sector area of ​​the virtual roulette wheel, start the rotation of the virtual roulette wheel, and when the rotation stops, delete the quantum bit corresponding to the sector area pointed by the virtual pointer from the virtual roulette wheel, repeat the above steps, and when the number of quantum bits on the virtual roulette wheel is equal to the quantity threshold, the quantum bits on the virtual roulette wheel are retained.

9. The electronic bidding system optimization method according to claim 8 is characterized in that: After the output end combines multiple groups of quantum bit sets to obtain a solution set, the quantum bit selection state in the solution set is automatically selected as the final selection solution output, including the following steps: The output end combines multiple groups of quantum bit sets to obtain a solution set, and compares the objective function values ​​of all quantum bits in the solution set in turn, and automatically selects the quantum bit selection state with the smallest objective function value as the final selection solution output.

10. The electronic bidding system optimization method according to claim 9, characterized in that: The processing end obtains all bidder information related to the bidding project and represents the selection status of each bidder as a quantum bit, including the following steps: The processing end obtains relevant information of all bidders from the bidding system, including the bidder's identification, bidder's quotation or cost, quality score of the product or service provided by the bidder, delivery time of the bidder, assessment of the bidder's performance capability, and whether the bidder meets the technical requirements of the project; The relevant information of the bidder is used as input data, and a quantum bit is allocated to each bidder to indicate whether the bidder is selected. Suppose there are N bidders, there will be N quantum bits, and the selection state of each quantum bit is expressed as: |0>: indicates that the bidder is not selected, |1>: indicates that the bidder is selected.

Images