{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# 利用PINNs求解二维带点源泊松方程\n", "\n", "[![下载Notebook](https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/website-images/master/resource/_static/logo_notebook.svg)](https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/master/mindflow/zh_cn/physics_driven/mindspore_poisson_point_source.ipynb) [![下载样例代码](https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/website-images/master/resource/_static/logo_download_code.svg)](https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/master/mindflow/zh_cn/physics_driven/mindspore_poisson_point_source.py) [![查看源文件](https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/website-images/master/resource/_static/logo_source.svg)](https://gitee.com/mindspore/docs/blob/master/docs/mindflow/docs/source_zh_cn/physics_driven/poisson_point_source.ipynb)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 问题描述\n", "\n", "本案例演示如何利用PINNs方法求解二维带点源的泊松方程:\n", "\n", "$$\n", "\\Delta u = - \\delta(x-x_{src})\\delta(y-y_{src}).\n", "$$\n", "\n", "其中 $(x_{src}, y_{src})$ 为点源位置对应的坐标。\n", "点源在数学上可以用狄拉克 $\\delta$ 函数来表示:\n", "\n", "$$\n", "\\delta(x) = \\begin{cases}\n", "+\\infty, & x = 0 \\\\\n", "0, & x \\neq 0\n", "\\end{cases}\n", "\\qquad\n", "\\int_{-\\infty}^{+\\infty}\\delta(x)dx = 1.\n", "$$\n", "\n", "当求解域 $\\Omega=[0,\\pi]^2$ 时,\n", "二维带点源的泊松方程的解析解为:\n", "\n", "$$\n", "u(x,y) = \\frac{4}{\\pi^2} \\sum_{i=1}^{\\infty} \\sum_{j=1}^{\\infty}\\frac{\\sin{(i x)}\\sin{(i x_{src})}\\sin{(j y)}\\sin{(j y_{src})}}{i^2 + j^2}.\n", "$$\n", "\n", "与该案例相对应的论文为:\n", "[Xiang Huang, Hongsheng Liu, Beiji Shi, Zidong Wang, Kang Yang, Yang Li, Min Wang, Haotian Chu, Jing Zhou, Fan Yu, Bei Hua, Bin Dong, Lei Chen. “A Universal PINNs Method for Solving Partial Differential Equations with a Point Source”. Thirty-First International Joint Conference on Artificial Intelligence (IJCAI 2022), Vienna, Austria, Jul, 2022, Pages 3839-3846.](https://www.ijcai.org/proceedings/2022/0533.pdf)\n", "\n", "## 技术路径\n", "\n", "MindSpore Flow求解该问题的具体流程如下:\n", "\n", "1. 创建训练数据集。\n", "2. 构建模型。\n", "3. 优化器。\n", "4. 约束。\n", "5. 模型训练。\n", "6. 模型推理及可视化。" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import time\n", "\n", "from mindspore import context, nn, ops, jit\n", "from mindflow import load_yaml_config\n", "\n", "from src.dataset import create_train_dataset, create_test_dataset\n", "from src.poisson import Poisson\n", "from src.utils import calculate_l2_error, visual\n", "\n", "context.set_context(mode=context.GRAPH_MODE, save_graphs=False, device_target=\"GPU\")\n", "\n", "# Load config\n", "file_cfg = \"poisson_cfg.yaml\"\n", "config = load_yaml_config(file_cfg)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 创建数据集\n", "\n", "本案例在求解域、边值条件、点源区域(以点源位置为中心的矩形区域)进行随机采样,生成训练数据集。具体方法见``src/dataset.py``。" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "# Create the dataset\n", "ds_train = create_train_dataset(config)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 构建模型\n", "\n", "本案例采用结合了sin激活函数的多尺度神经网络。" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "from mindflow.cell import MultiScaleFCSequential\n", "\n", "# Create the model\n", "model = MultiScaleFCSequential(config['model']['in_channels'],\n", " config['model']['out_channels'],\n", " config['model']['layers'],\n", " config['model']['neurons'],\n", " residual=True,\n", " act=config['model']['activation'],\n", " num_scales=config['model']['num_scales'],\n", " amp_factor=1.0,\n", " scale_factor=2.0,\n", " input_scale=[10., 10.],\n", " )" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 约束\n", "\n", "在利用``mindflow``求解PDE时,我们需要写一个``mindflow.PDEWithLloss``的子类来定义控制方程和边界条件分别对应的损失函数项(``loss_pde``和``loss_bc``)。因为点源区域需要加密采样点,所以我们额外增加了一个损失函数项(``loss_src``)。\n", "\n", "当PINNs方法将控制方程的残差作为损失函数项来约束神经网络时,狄拉克$\\delta$函数的奇异性使得神经网络的训练无法收敛,因此我们采用二维拉普拉斯分布的概率密度函数去近似狄拉克 $\\delta$ 函数,即:\n", "\n", "$$\n", "\\eta_{\\alpha}(x, y) = \\frac{1}{4\\alpha^2} exp({-\\frac{|x-x_{src}|+|y-y_{src}|}{\\alpha}}) \\qquad \\underrightarrow{approx} \\qquad \\delta(x-x_{src})\\delta(y-y_{src}).\n", "$$\n", "\n", "其中 $\\alpha$ 为核宽度。理论上来说,只要核宽度 $\\alpha$ 充分小,那么上述概率密度函数就能很好地近似狄拉克 $\\delta$ 函数。但是实际上核宽度 $\\alpha$ 的选取对于近似效果有着重要影响。当 $\\alpha$ 太大时,概率密度函数 $\\eta_{\\alpha}(x, y)$ 与狄拉克 $\\delta$ 函数之间的近似误差会变大。但如果 $\\alpha$ 太小,训练过程可能不会收敛,或者收敛后的精度可能很差。因此,$\\alpha$ 需要进行手工调参。我们这里将其确定为 $\\alpha=0.01$。\n", "\n", "在求解区域内、边界、点源区域均采用L2损失,并利用``mindflow``的``MTLWeightedLoss``多目标损失函数将这三个损失函数项结合起来。" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "poisson: Derivative(u(x, y), (x, 2)) + Derivative(u(x, y), (y, 2)) + 2500.0*exp(-100.0*Abs(x - pi/2))*exp(-100.0*Abs(y - pi/2))\n", " Item numbers of current derivative formula nodes: 3\n", "bc: u(x, y)\n", " Item numbers of current derivative formula nodes: 1\n" ] } ], "source": [ "import sympy\n", "from mindspore import numpy as ms_np\n", "from mindflow import PDEWithLoss, MTLWeightedLoss, sympy_to_mindspore\n", "\n", "class Poisson(PDEWithLoss):\n", " \"\"\"Define the loss of the Poisson equation.\"\"\"\n", "\n", " def __init__(self, model):\n", " self.x, self.y = sympy.symbols(\"x y\")\n", " self.u = sympy.Function(\"u\")(self.x, self.y)\n", " self.in_vars = [self.x, self.y]\n", " self.out_vars = [self.u,]\n", " self.alpha = 0.01 # kernel width\n", " super(Poisson, self).__init__(model, self.in_vars, self.out_vars)\n", " self.bc_nodes = sympy_to_mindspore(self.bc(), self.in_vars, self.out_vars)\n", " self.loss_fn = MTLWeightedLoss(num_losses=3)\n", "\n", " def pde(self):\n", " \"\"\"Define the gonvering equation.\"\"\"\n", " uu_xx = sympy.diff(self.u, (self.x, 2))\n", " uu_yy = sympy.diff(self.u, (self.y, 2))\n", "\n", " # Use Laplace probability density function to approximate the Dirac \\delta function.\n", " x_src = sympy.pi / 2\n", " y_src = sympy.pi / 2\n", " force_term = 0.25 / self.alpha**2 * sympy.exp(-(\n", " sympy.Abs(self.x - x_src) + sympy.Abs(self.y - y_src)) / self.alpha)\n", "\n", " poisson = uu_xx + uu_yy + force_term\n", " equations = {\"poisson\": poisson}\n", " return equations\n", "\n", " def bc(self):\n", " \"\"\"Define the boundary condition.\"\"\"\n", " bc_eq = self.u\n", "\n", " equations = {\"bc\": bc_eq}\n", " return equations\n", "\n", " def get_loss(self, pde_data, bc_data, src_data):\n", " \"\"\"Define the loss function.\"\"\"\n", " res_pde = self.parse_node(self.pde_nodes, inputs=pde_data)\n", " res_bc = self.parse_node(self.bc_nodes, inputs=bc_data)\n", " res_src = self.parse_node(self.pde_nodes, inputs=src_data)\n", "\n", " loss_pde = ms_np.mean(ms_np.square(res_pde[0]))\n", " loss_bc = ms_np.mean(ms_np.square(res_bc[0]))\n", " loss_src = ms_np.mean(ms_np.square(res_src[0]))\n", "\n", " return self.loss_fn((loss_pde, loss_bc, loss_src))\n", "\n", "# Create the problem and optimizer\n", "problem = Poisson(model)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 优化器\n", "\n", "本案例采用Adam优化器,并在训练进行到40%、60%、80%时,学习率衰减为初始学习率的1/10、1/100、1/1000。" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "n_epochs = 250\n", "\n", "params = model.trainable_params() + problem.loss_fn.trainable_params()\n", "steps_per_epoch = ds_train.get_dataset_size()\n", "milestone = [int(steps_per_epoch * n_epochs * x) for x in [0.4, 0.6, 0.8]]\n", "lr_init = config[\"optimizer\"][\"initial_lr\"]\n", "learning_rates = [lr_init * (0.1**x) for x in [0, 1, 2]]\n", "lr_ = nn.piecewise_constant_lr(milestone, learning_rates)\n", "optimizer = nn.Adam(params, learning_rate=lr_)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 模型训练\n", "\n", "使用MindSpore>= 2.0.0的版本,可以使用函数式编程范式训练神经网络。" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "epoch: 1 train loss: 27463.34765625 epoch time: 21.84s\n", "epoch: 2 train loss: 22351.10351562 epoch time: 12.35s\n", "epoch: 3 train loss: 15621.81347656 epoch time: 12.36s\n", "epoch: 4 train loss: 11910.88671875 epoch time: 12.41s\n", "epoch: 5 train loss: 7340.56933594 epoch time: 12.45s\n", "epoch: 6 train loss: 7032.89843750 epoch time: 12.44s\n", "epoch: 7 train loss: 3993.34130859 epoch time: 12.46s\n", "epoch: 8 train loss: 2885.75390625 epoch time: 12.43s\n", "epoch: 9 train loss: 1879.61999512 epoch time: 12.40s\n", "epoch: 10 train loss: 3002.69677734 epoch time: 12.36s\n", "...\n", "epoch: 241 train loss: 6.00261974 epoch time: 12.34s\n", "epoch: 242 train loss: 6.08083344 epoch time: 12.38s\n", "epoch: 243 train loss: 6.05793428 epoch time: 12.36s\n", "epoch: 244 train loss: 6.05963135 epoch time: 12.36s\n", "epoch: 245 train loss: 5.88465643 epoch time: 12.43s\n", "epoch: 246 train loss: 6.06778002 epoch time: 12.36s\n", "epoch: 247 train loss: 5.94315052 epoch time: 12.40s\n", "epoch: 248 train loss: 6.28999186 epoch time: 12.44s\n", "epoch: 249 train loss: 5.82442093 epoch time: 12.47s\n", "epoch: 250 train loss: 5.86588335 epoch time: 12.44s\n", "End-to-End total time: 3099.2 s\n" ] } ], "source": [ "def train():\n", " grad_fn = ops.value_and_grad(problem.get_loss, None, optimizer.parameters, has_aux=False)\n", "\n", " @jit\n", " def train_step(pde_data, bc_data, src_data):\n", " loss, grads = grad_fn(pde_data, bc_data, src_data)\n", " if use_ascend:\n", " loss = loss_scaler.unscale(loss)\n", " is_finite = all_finite(grads)\n", " if is_finite:\n", " grads = loss_scaler.unscale(grads)\n", " loss = ops.depend(loss, optimizer(grads))\n", " loss_scaler.adjust(is_finite)\n", " else:\n", " loss = ops.depend(loss, optimizer(grads))\n", " return loss\n", "\n", " def train_epoch(model, dataset, i_epoch):\n", " local_time_beg = time.time()\n", "\n", " model.set_train()\n", " for _, (pde_data, bc_data, src_data) in enumerate(dataset):\n", " loss = train_step(pde_data, bc_data, src_data)\n", "\n", " print(\n", " f\"epoch: {i_epoch} train loss: {float(loss):.8f}\" +\n", " f\" epoch time: {time.time() - local_time_beg:.2f}s\")\n", "\n", " for i_epoch in range(1, 1 + n_epochs):\n", " train_epoch(model, ds_train, i_epoch)\n", "\n", "time_beg = time.time()\n", "train()\n", "print(f\"End-to-End total time: {time.time() - time_beg:.1f} s\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 模型推理及可视化\n", "\n", "计算相对L2误差以及绘制参考解和模型预测结果的对比图。" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Relative L2 error: 0.0154\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from src.utils import calculate_l2_error, visual\n", "\n", "# Create the dataset\n", "ds_test = create_test_dataset(config)\n", "\n", "# Evaluate the model\n", "calculate_l2_error(model, ds_test)\n", "\n", "# Visual comparison of label and prediction\n", "visual(model, ds_test)" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3.8.16 ('ms')", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.16" }, "vscode": { "interpreter": { "hash": "c22ff8496cdfc43b41d028d0afe27e7d77fc6967d8e63387d8409afb66bbd90b" } } }, "nbformat": 4, "nbformat_minor": 2 }