User multi-dimensional bounded rationality energy consumption behavior model description and parameter identification method

Abstract

The application discloses a user multi-dimensional bounded rationality energy consumption behavior model depiction and parameter identification method, and steps are as follows: a power user multi-element energy demand framework containing multi-dimensional demand is established; energy consumption terminal equipment corresponding to the multi-element demand is divided into time-sensitive type, temperature-sensitive type and power-sensitive type, and physical characteristic models thereof are respectively established; a user bounded rationality energy multi-dimensional prospect function facing multi-element demand is constructed; a setting method of reference points of satisfaction degrees of various demands and a declarative preference of the user to various power demands are determined; a logarithmic likelihood function of a risk coefficient, a loss-aversion coefficient in a value function of the prospect function and a personal preference coefficient in a weight function is constructed; parameters of the prospect function are identified, a parameter distribution conforming to actual selection preferences of the user is obtained, and the bounded rationality energy consumption behavior model of the user is determined. The application can effectively depict bounded rationality energy decision-making behaviors of the user in actual production and life.

Description

Technical Field

[0001] This invention relates to a method for characterizing and modeling demand-side energy consumption behavior, and in particular to a method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model. Background Technology

[0002] With the rapid growth of demand-side resources and the development of measurement and control technologies, demand-side resources have become crucial flexibility adjustment resources for future new energy systems. However, energy users have limited cognitive abilities, willpower, and decision-making calculation capabilities, coupled with subjective preferences in their decision-making, leading to bounded rationality in their energy consumption behavior. Currently, research based on the traditional classical economic "economic man" assumption mostly aims to minimize energy costs or maximize utility in characterizing energy consumption behavior. However, such simplistic and idealized characterizations of energy consumption behavior often fail to accurately describe and explain the impact of multiple, time-varying, and interrelated factors on user behavior, resulting in modeling results that do not match reality and overly optimistic assessments of load regulation capacity or flexibility. Therefore, it is necessary to introduce behavioral economics theory with the bounded rationality of "natural man" as a basic assumption. Starting from the diverse energy needs of actual users, this approach can reasonably characterize users' bounded rational energy consumption behavior, providing a fundamental theoretical basis and technical methodological support for research on market mechanisms such as energy retail, end-to-end trading, and demand response, as well as the assessment and regulation of demand-side flexibility resources in power systems. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to provide a method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model, thereby enabling a practical and reasonable characterization of the user's bounded rationality energy consumption behavior model and parameter identification.

[0004] Technical solution: The present invention provides a method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model, comprising the following steps:

[0005] (1) Establish a diversified energy demand framework for electricity users based on mental accounting theory.

[0006] (2) Divide the energy-consuming terminal equipment corresponding to the diverse needs into three categories: time-sensitive, temperature-sensitive and power-sensitive, and establish their physical characteristic models respectively.

[0007] (3) Construct a multidimensional prospect function for user bounded rational energy use based on cumulative prospect theory to meet diverse needs.

[0008] (4) Based on historical load data and social survey data, determine the method for setting reference points for various demand satisfaction and users’ declarative preferences for various electricity demands.

[0009] (5) Based on the hybrid Logit model, construct a probability model for users’ bounded rational energy use decision, and establish the log-likelihood functions of the risk coefficient, loss aversion coefficient and personal preference coefficient in the value function of the prospect function and the weight function.

[0010] (6) The maximum simulation likelihood estimation method is used to identify the parameters of the prospect function, obtain the parameter distribution that conforms to the user's actual choice preference, and clarify the user's bounded rationality energy consumption behavior model.

[0011] The specific steps (1) are as follows:

[0012] (1.1) Establish the basic part of the diversified energy demand architecture for electricity users. The basic architecture includes food demand (corresponding to the use demand of equipment such as induction cookers, refrigerators, and rice cookers), temperature demand (corresponding to the use demand of equipment such as air conditioners and heaters), hygiene demand (corresponding to the power demand of equipment such as washing machines, vacuum cleaners, and hair dryers), lighting demand (corresponding to the power demand of lamps), travel demand (corresponding to the charging demand of electric vehicles and electric vehicles), and entertainment / work demand (corresponding to the power demand of equipment such as computers, mobile phones, televisions, and treadmills). Each demand belongs to a different mental account and is independent of each other.

[0013] (1.2) Establish the upper part of the diversified energy demand architecture for electricity users. The upper part of the architecture includes social demand and electricity cost demand. The social demand includes herd mentality and environmental awareness. Some users may consider how to save electricity bills while meeting basic needs, while others will spontaneously respond to the call of policies.

[0014] Step (2) specifically involves:

[0015] (2.1) Model the physical characteristics of time-sensitive devices.

[0016] For time-sensitive devices (a), which mainly include rice cookers, dishwashers, washing machines, vacuum cleaners, and water heaters, their power consumption model is represented by the following formula:

[0017]

[0018] In the formula, q a The daily electricity consumption of device a; The power of the device during time period t; This indicates the start / stop status of the equipment during this time period; and These represent the upper and lower limits of the time range during which device a can be moved, respectively.

[0019] The following is a calculation of residential users' comfort level with device a, based on the device's time offset ratio:

[0020]

[0021] In the formula, u a For comfort level; T a0 and T a These are the initial usage times of the equipment and the usage times after its transfer.

[0022] (2.2) Model the physical characteristics of temperature-sensitive equipment.

[0023] Temperature-sensitive devices include appliances such as air conditioners and refrigerators. Their load power models are constructed using thermodynamic equivalent models.

[0024]

[0025] In the formula, θ(t) and θ a (t) represents the indoor and outdoor temperatures at time t, respectively, in °C; C is the equivalent heat capacity of the temperature-sensitive equipment, in kW·h / °C; R is the equivalent thermal resistance, in °C / kW; P c The cooling / heating power of the equipment is expressed in kW; the cooling / heating power and the electrical power consumption P of the equipment satisfy a certain proportional relationship, P c =ηP, where η is the energy efficiency ratio; m(t) represents the equipment's on / off state, taking a value of 0 when the equipment is off and 1 when it is on; ε represents the simulation step size; θ - and θ + Indicates the upper and lower limits of indoor temperature.

[0026] Taking the air conditioner, a temperature-controlled load that accounts for the largest proportion of the electrical load, as the target, its load power model is expressed by the following formula:

[0027]

[0028] In the formula, P a (t) represents the operating power of the air conditioner at time t; T represents the upper limit of operating power. in (t), T out (t) represents the indoor and outdoor temperatures at time t, respectively; ε is the inertia coefficient for the change in indoor temperature; η a A is the thermal conductivity; T is the thermal conductivity. set (t) represents the air conditioning set temperature at time t; ΔT represents the maximum temperature offset.

[0029] (2.3) Model the physical characteristics of power-sensitive devices.

[0030] Power-sensitive devices mainly include electric vehicles, mobile phones, and computers. Taking high-power electric vehicles as an example, considering the relationship between the electric vehicle's state of charge and charging power, charging power constraints, and state of charge constraints, its load power model is expressed by the following formula:

[0031]

[0032] In the formula, SOC(t) represents the state of charge of the electric vehicle at time t; P c (t) represents the charging power of the electric vehicle at time t; η c E represents charging efficiency; E represents the rated capacity of the electric vehicle battery. Maximum charging power for electric vehicles; SOC max SOC min These represent the maximum and minimum states of charge of an electric vehicle, respectively.

[0033] Step (3) specifically involves:

[0034] (3.1) Calculate the prospect theory value function, the formula is as follows:

[0035]

[0036]

[0037] In the formula, v + (x) and v - (x) represents the value function of positively and negatively correlated indicators, respectively; x is the attribute indicator value of user satisfaction with electricity consumption to meet multidimensional user needs, such as the user's waiting time for dining, laundry, and physical comfort; x0 is the corresponding reference point; α and β are the risk preference coefficient and risk aversion coefficient, respectively; λ is the decision-maker's sensitivity coefficient to loss and gain; W is the indicator weight.

[0038] (3.2) Calculate the weight function based on the cumulative prospect theory.

[0039] Suppose that a certain attribute of a user's requirement has a result (x1 < ... < x l-1 <x l <...<x n The corresponding probability is (p1,…,p). l-1 ,p l ,…,p n ), where x l-1 Representing loss, x l If the result represents the weighting function, then the weighting function is as follows:

[0040]

[0041] ω(p1)=π(p1)

[0042] ω(p n )=π(p n )

[0043]

[0044] In the formula, π(p) is the initial weight function; ω(p) is the cumulative weight function; and γ is the risk attitude coefficient of the decision-maker towards gains and losses.

[0045] (3.3) For any electricity consumption scheme with multidimensional electricity demand, its prospect function is as follows:

[0046]

[0047] H={α j ,β j ,λ j ,γ j |j=1,2,...,m}

[0048] W = {W j |j=1,2,...,m}

[0049]

[0050]

[0051] X 0i ={x 0ij |j=1,2,...,m}

[0052] In the formula, V i () represents the comprehensive prospect value of electricity consumption scheme i; H is the parameter vector of each attribute; W is the index weight vector; X i and P i X represents the vector of each index value and its corresponding probability for scheme i; 0i U is the reference point vector; i ε is a definite term in the prospect function. i This is the error term.

[0053] Step (4) specifically involves:

[0054] By combining historical load data of various electrical devices used by users, the cost and comfort index values ​​and corresponding probabilities of each user's electricity consumption mode are calculated, and represented by vector X. i and P i This indicates that, based on the user's initial electricity consumption pattern, reasonable decision reference points are set to meet various production and living needs, including cost and electricity comfort, using vector X. 0i express.

[0055] The user-based bounded rationality energy consumption decision probability model constructed in step (5) based on the hybrid Logit model is as follows:

[0056]

[0057]

[0058]

[0059] In the formula, prob ig The probability of user g choosing option i; β is the set of all parameters to be estimated; f(β|θ) is the probability density function of the parameters; θ is the unknown characteristic parameter of the probability density function of the parameters; k is the number of options; For scheme i, the determined part of the foreground value for user g.

[0060] Step (6) specifically involves:

[0061] (6.1) For a dataset containing N users, its log-likelihood function is:

[0062]

[0063] Y = {y gq |q=1,2,...,Q;g=1,2,...,N}

[0064] In the formula, N represents the number of users; Q represents the number of decision scenarios; k represents the number of options; Y represents the user's decision outcome matrix; y gq Let y represent the decision made by user g in scenario q, and y represent the decision made by user g in scenario q. gq For any given ∈{1,2,...,k}, if the user selects scheme i, then the expression y... gq = The value of i is 1, otherwise it is 0.

[0065] (6.2) Calculate the simulated probability. Given θ, a random vector is drawn from the probability density function f(β|θ) by Latin hypercube sampling, denoted as β. 1 As a preference parameter for user 1, calculate the simulated probability value of user 1 choosing any option, and repeat the above operation N times.

[0066] (6.3) The optimal θ value is obtained by using a genetic algorithm so that the maximum likelihood operator is maximized and the parameters of the multidimensional prospect function used to characterize the energy consumption behavior of bounded rationality are clarified.

[0067] A computer storage medium storing a computer program, which, when executed by a processor, implements the aforementioned method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model.

[0068] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model.

[0069] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0070] 1. It can characterize the bounded rationality of users' energy consumption behavior based on actual production and life, and provide basic theoretical and methodological support for the design of energy demand-side market mechanisms and the research on the flexible exploration, assessment and regulation of resources;

[0071] 2. It can comprehensively utilize actual energy load data and social survey data on energy consumption behavior to identify parameters of bounded rationality behavior models for users, thereby improving the practicality and accuracy of bounded rationality behavior characterization models and methods. Attached Figure Description

[0072] Figure 1 This is a schematic diagram of the method described in this invention;

[0073] Figure 2 A schematic diagram illustrating the architecture for users' diverse energy needs;

[0074] Figure 3 A schematic diagram of the cumulative prospect theory value function;

[0075] Figure 4 This is a schematic diagram of the cumulative prospect theory weighting function. Detailed Implementation

[0076] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0077] Example 1:

[0078] like Figure 1 As shown, the method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior includes the following steps:

[0079] Step 1: Establish a framework for users' diverse energy needs.

[0080] Users' energy consumption needs can be divided into multiple dimensions, such as Figure 2As shown. The most basic needs are various daily life requirements, which are the fundamental driving force behind residential electricity consumption. These mainly include: food needs (corresponding to the use of appliances such as induction cookers, refrigerators, and rice cookers), temperature needs (corresponding to the use of appliances such as air conditioners and heaters), hygiene needs (corresponding to the electricity needs of appliances such as washing machines, vacuum cleaners, and hair dryers), lighting needs (corresponding to the electricity needs of lamps), travel needs (corresponding to the charging needs of electric vehicles and electric bicycles), and entertainment / work needs (corresponding to the electricity needs of devices such as computers, mobile phones, televisions, and treadmills). Each need belongs to a different mental account and is independent of the others. In addition, energy consumption behavior is also influenced by social needs such as conformity and environmental awareness, as well as electricity cost considerations. Some users may consider how to save on electricity bills while meeting basic needs, while others will spontaneously respond to policy calls. These two constitute the second level of user needs, namely, social attribute needs.

[0081] Step 2: Perform physical characteristic modeling on the terminal device.

[0082] For time-sensitive devices (mainly including rice cookers, dishwashers, washing machines, vacuum cleaners, etc.), their power consumption model can be expressed by the following formula:

[0083]

[0084] Where, q a The daily electricity consumption of device a; The power of the device during time period t; This indicates the start / stop status of the equipment during this time period; and These represent the upper and lower limits of the time range during which device a can be moved, respectively.

[0085] The following is a calculation of residential users' comfort level with device a, based on the device's time offset ratio:

[0086]

[0087] Among them, u a For comfort level; T a0 and T a These are the initial usage times of the equipment and the usage times after its transfer.

[0088] For temperature-sensitive equipment, taking air conditioning as an example, its load power model can be expressed by the following formula:

[0089]

[0090] Among them, P a (t) represents the operating power of the air conditioner at time t; T represents the upper limit of operating power.in (t), T out (t) represents the indoor and outdoor temperatures at time t, respectively; ε is the inertia coefficient for the change in indoor temperature; η a A is the thermal conductivity; T is the thermal conductivity. set (t) represents the air conditioning set temperature at time t; ΔT represents the maximum temperature offset.

[0091] For power-sensitive devices, taking electric vehicles as the target, their load power model is expressed by the following formula:

[0092]

[0093] Where SOC(t) is the state of charge of the electric vehicle at time t; P c (t) represents the charging power of the electric vehicle at time t; η c E represents charging efficiency; E represents the rated capacity of the electric vehicle battery. Maximum charging power for electric vehicles; SOC max SOC min These represent the maximum and minimum states of charge of an electric vehicle, respectively.

[0094] Step 3: Construct a multidimensional prospect function for the user's bounded rational energy use.

[0095] The formula for calculating the prospect theory value function is as follows:

[0096]

[0097]

[0098] Among them, v + (x) and v - (x) represents the value function of positively and negatively correlated indicators, respectively; x is the attribute indicator value of user satisfaction with electricity consumption to meet multidimensional user needs, such as the user's waiting time for dining, laundry, and physical comfort; x0 is the corresponding reference point; α and β are the risk preference coefficient and risk aversion coefficient, respectively; λ is the decision-maker's sensitivity coefficient to loss and gain; W is the indicator weight.

[0099] Suppose that a certain attribute of a user's requirement has a result (x1 < ... < x l-1 <x l <...<x n The corresponding probability is (p1,…,p). l-1 ,p l ,…,p n ), where x l-1 Representing loss, x l If the result represents the weighting function, then the weighting function is as follows:

[0100]

[0101] Where π(p) is the initial weighting function; ω(p) is the cumulative weighting function; and γ is the risk attitude coefficient of the decision-maker towards gains and losses.

[0102] For any electricity consumption scheme with multidimensional electricity demand, its prospect function is as follows:

[0103]

[0104] Among them, V i () represents the comprehensive prospect value of electricity consumption scheme i; H is the parameter vector of each attribute; W is the index weight vector; X i and P i X represents the vector of each index value and its corresponding probability for scheme i; 0i U is the reference point vector; i ε is a definite term in the prospect function. i This is the error term.

[0105] Step 4: Calculate the index value X for each user's electricity consumption plan based on historical load data. i and P i And set reference point X according to the user's initial power consumption pattern. 0i .

[0106] Step 5: Based on the hybrid Logit model, construct the user selection probability model as follows:

[0107]

[0108]

[0109]

[0110] Among them, prob ig The probability of user g choosing option i; β is the set of all parameters to be estimated; f(β|θ) is the probability density function of the parameters; θ is the unknown characteristic parameter of the probability density function of the parameters; k is the number of options; For scheme i, the determined part of the foreground value for user g.

[0111] Step Six: Identify the parameters of the prospect function using the maximum simulated likelihood estimation method to obtain a parameter distribution that matches the user's actual selection preferences, as detailed below:

[0112] First, we construct the log-likelihood operator. For a dataset containing N users, its log-likelihood function is:

[0113]

[0114] Where N is the number of users; Q is the number of decision scenarios; k is the number of solutions; Y represents the user's decision result matrix; y gq Let y represent the decision made by user g in scenario q, and y represent the decision made by user g in scenario q. gq For any given ∈{1,2,...,k}, if the user selects scheme i, then the expression y... gq = The value of i is 1, otherwise it is 0.

[0115] Next, the simulated probability is calculated. Given θ, a random vector, denoted as β, is drawn from the probability density function f(β|θ) using Latin hypercube sampling. 1 As a preference parameter for user 1, calculate the simulated probability value of user 1 choosing any option, and repeat the above operation N times.

[0116] Finally, a genetic algorithm is used to solve for the optimal θ value, so that the simulated maximum likelihood operator reaches its maximum value, the parameters of the multidimensional prospect function are obtained, and the model characterizing bounded rational behavior is clarified.

[0117] Example 2:

[0118] To verify the feasibility and correctness of this invention, the following example is provided:

[0119] This example uses daily load data from 2,858 residential users and social survey questionnaire data as a basis to identify parameters based on users' bounded rational electricity consumption behavior for time-sensitive devices (rice cookers).

[0120] Table 1. Partial Statement of Preferences Questionnaire

[0121]

[0122] Based on the above data, a user bounded rationality energy consumption prospect function is constructed according to the modeling method described above. The user choice probability is calculated using a mixed Logit model, and then the optimal parameter distribution is obtained through maximum simulation likelihood estimation. Three different models are used to verify the applicability of the constructed model, employing the Akaike Information Criterion (AIC) and the McFadden pseudo-R². 2 Evaluating model fit: The smaller the AIC value, the lower the pseudo-R. 2 The larger the value, the better the model fit. The results are shown in Table 2.

[0123] (1) Model 1: Combination of traditional random utility model and hybrid Logit model.

[0124] (2) Model 2: Combining cumulative prospect theory with the hybrid Logit model, but using the same preference parameters for all attributes.

[0125] (3) Model 3: The cumulative prospect theory is combined with the hybrid Logit model, and the parameter differences of different attributes are considered.

[0126] Table 2 Parameter Identification Results

[0127]

[0128] The parameter identification results show that Model 3, the model described in this invention, has the best fit, and the obtained parameter distribution best reflects the electricity consumption decision preferences of residential users in real life, providing theoretical and parameter support for the formulation of demand response strategies.

Claims

1. A method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model, characterized in that, Includes the following steps: (1) Establish a diversified energy demand framework for electricity users based on mental accounting theory; (2) Divide the energy-consuming terminal equipment corresponding to the diverse needs into three categories: time-sensitive, temperature-sensitive, and power-sensitive, and establish their physical characteristic models respectively; (3) Construct a multidimensional prospect function for user-bounded rational energy use that addresses diverse needs based on cumulative prospect theory; (4) Based on historical load data and social survey data, determine the method for setting reference points for various demand satisfaction and the users' declarative preferences for various electricity demands; (5) Based on the hybrid Logit model, construct a probability model for users' bounded rationality energy use decisions, and establish the log-likelihood functions of the risk coefficient, loss aversion coefficient and personal preference coefficient in the value function of the prospect function and the weight function. (6) The maximum simulation likelihood estimation method is used to identify the parameters of the prospect function, obtain the parameter distribution that conforms to the user's actual choice preference, and clarify the user's bounded rationality energy consumption behavior model; Step (2) is as follows: (2.1) Model the physical characteristics of time-sensitive devices; For time-sensitive devices This mainly includes rice cookers, dishwashers, washing machines, vacuum cleaners, and water heaters, whose power consumption model is expressed by the following formula: In the formula, For equipment Daily electricity consumption; For the equipment during the time period The power; This indicates the start / stop status of the equipment during this time period; and respectively equipment The upper and lower limits of the movable time range; Residential users' relationships with devices are constructed using the device time offset ratio. The user comfort level is as follows: In the formula, This represents the comfort level. and These are the initial usage times of the equipment and the usage times after its transfer; (2.2) Model the physical characteristics of temperature-sensitive equipment; Temperature-sensitive devices, including air conditioners and refrigerators, have their load power models constructed using thermodynamic equivalent models: In the formula, and They are respectively The indoor and outdoor temperatures at any given time. ; For the equivalent heat capacity of temperature-sensitive equipment, ; For equivalent thermal resistance, ; The cooling / heating capacity of the equipment. Cooling / heating power and equipment power consumption To satisfy a certain proportional relationship, , Energy efficiency ratio; This indicates the on / off status of the device; the value is 0 when the device is off and 1 when the device is on. Indicates the simulation step size; and Indicates the upper and lower limits of indoor temperature; (2.3) Perform physical characteristic modeling for power-sensitive devices; Power-sensitive devices mainly include electric vehicles, mobile phones, and computers, and their load power model is expressed by the following formula: In the formula, The state of charge of the power-sensitive device at time t; The charging power of the device at time t; E represents the charging efficiency; E represents the rated capacity of the device's battery. The maximum charging power of the device; , These represent the maximum and minimum states of charge of the equipment, respectively.

2. The method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior according to claim 1, characterized in that, The specific steps (1) are as follows: (1.1) Establish the basic part of the diversified energy demand architecture of electricity users. The basic architecture includes food demand, temperature demand, hygiene demand, lighting demand, travel demand, and entertainment / work demand. Each demand belongs to a different mental account and is independent of each other. (1.2) Establish the upper part of the diversified energy demand architecture of electricity users, the upper architecture including social demand and electricity cost demand, the social demand including herd mentality and environmental awareness.

3. The method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior according to claim 1, characterized in that, The specific steps (3) are as follows: (3.1) Calculate the prospect theory value function, as shown in the following formula: In the formula, and These represent the value functions for positively and negatively correlated indicators, respectively. To meet users' multidimensional needs, the attribute index values ​​for electricity satisfaction; For the corresponding reference point; and These are the risk preference coefficient and the risk aversion coefficient, respectively. This represents the decision-maker's sensitivity to losses and gains. As the indicator weight; (3.2) Calculate the weighting function based on cumulative prospect theory; Suppose that a certain attribute of a user's requirement has a result. The corresponding probability is ,in Represents loss, If the result represents the weighting function, then the weighting function is as follows: In the formula, This is the initial weight function; This is the cumulative weight function; The risk attitude coefficient of decision-makers towards gains and losses; (3.3) For any electricity consumption scheme with multidimensional electricity demand, its prospect function is as follows: In the formula, For power supply plan The overall prospects value; It is a parameter vector for each attribute; This is the indicator weight vector; and The respective schemes A vector of each index value and its corresponding probability; The reference point vector; It is a definite term in the prospect function; This is the error term.

4. The method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model according to claim 1, characterized in that, Step (4) specifically involves: By combining historical load data of various user electrical devices, the cost and comfort index values ​​and corresponding probabilities of each user's electricity consumption mode are calculated, and vector values ​​are used respectively. and This indicates that, based on users' initial electricity consumption patterns, reasonable decision-making reference points should be set to meet various production and living needs, including cost and electricity comfort, and the usage vector should be considered. express.

5. The method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model according to claim 1, characterized in that, The user-based bounded rationality energy use decision probability model constructed in step (5) based on the hybrid Logit model is as follows: In the formula, For users Choose a solution The probability of; The set of all parameters to be estimated; The probability density function with parameters; These are the unknown feature parameters of the parametric probability density function; The number of solutions; For the plan For users The definite part of the foreground value.

6. The method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model according to claim 1, characterized in that, The specific steps (6) are as follows: (6.1) For a given set of elements containing The dataset of n users has the following log-likelihood function: In the formula, For the number of users; The number of decision scenarios; The number of solutions; This represents the matrix of user decision outcomes; Indicates user In the context The decision made below, and If the user selects the plan Then the expression The value is 1 if it is 1, otherwise it is 0. (6.2) Calculate the simulation probability, given Under the premise of Latin hypercube sampling, from the probability density function We draw a random vector from the vector, denoted as . Using this as a preference parameter for User 1, calculate the simulated probability value of User 1 choosing any option, and repeat the above operation. Second-rate; (6.3) The optimal solution is obtained by using a genetic algorithm. The value is such that the maximum likelihood operator in the simulation reaches its maximum value, and the parameters of the multidimensional prospect function are clearly defined to characterize the bounded rational energy consumption behavior.

7. A computer storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements a method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model as described in any one of claims 1-6.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements a method for characterizing and identifying parameters of a user's multidimensional bounded rationality energy consumption behavior model as described in any one of claims 1-6.

Images